INTRODUCTION

The way in which chronic hepatitis is classified is undergoing re-evaluation. There are a number of reasons for this. They include an improved appreciation of the wide range of the possible aetiologies coupled with a more critical appreciation of the pathological changes seen. To a large this re-evaluation has been stimulated by the clinical and pathological features of hepatitis C.

DEFINITION OF CHRONICITY

Hepatitis is classified as chronic when there are abnormal liver function tests for over 6 months. This is to exclude cases of acute hepatitis which are slow to resolve. Histopathologists should be careful of labelling a case as one of chronic hepatitis in the absence of an appropriate clinical history unless unequivocal pathological features, e.g. ground glass cells in the case of HBV, are present. It should however be noted that auto-immune hepatitis, because of its particular natural history, is considered chronic even when the history is shorter than 6 months.

AETIOLOGICAL CLASSIFICATION

Aetiology is central to the classification of chronic hepatitis as it is the main determinant of treatment. Although many different liver diseases can cause a prolonged elevation of liver enzymes, by convention they are not all classified under the heading of chronic hepatitis. It has recently been agreed that while chronic hepatitis includes liver disease caused by viruses, drugs and auto-immune mechanisms it excludes metabolic diseases (e.g. WilsonΆs disease) and chronic biliary tract diseases (e.g. PBC). It should be noted that both of the later groups of conditions may produce clinical and histological changes very similar to those seen in chronic hepatitis and are excluded only by convention. They frequently pose problems in differential diagnosis for the histopathologist considering a possible case of chronic hepatitis. The range of hepatotrophic viruses capable of causing chronic liver disease has continued to expand. They now include: HBV, HCV, HDV and HGV. Their diagnosis is almost always based on serology (including PCR). It is increasingly appreciated that an individual may be infected by more than one of these viruses and that this influences the natural history. Auto-immune hepatitis, which is also diagnosed serologically, can also be subdivided into a number of subtypes which have different clinical and serological features (as well as responses to treatment). Drugs, can of course, cause virtually every pattern of liver disease and this includes chronic hepatitis. A careful history is needed in all problematic cases.

Even this subdivision is an over simplification. For example chronic HDV may be associated with auto-immune features and some cases of drug induced hepatitis. e.g. those caused by tienilic acid, may also be associated with auto-antibodies. Furthermore co-infection with HIV may alter the natural history of infection by the hepatotrophic viruses as may other co-existing causes of liver disease e.g. alcohol. It is worth noting that alcohol rarely, if ever, causes a chronic hepatitis. Cases previously ascribed to alcohol are now known to be virtually all caused by HCV. Finally there are some cases of chronic hepatitis for which no cause, at present, can be identified.

HISTOLOGICAL CLASSIFICATION

In cases of chronic hepatitis there may be a wide range of inflammatory and/or degenerative changes. These include inflammation confined to the portal tracts (sometimes in the form of lymphoid follicles), piecemeal necrosis («interface hepatitis»), lobular inflammation and degenerative changes (including fatty change and acidophilic degeneration). Similarly fibrosis may range from portal tract expansion to portal - portal linking to portal - venous bridging to cirrhosis. Clearly it is desirable to be able to incorporate the full spectrum of histological changes in an ideal scoring system.

Until recently chronic hepatitis has been classified into chronic persistent (cph) and chronic active hepatitis (cah) based solely on the presence or absence of piecemeal necrosis. The presence or absence of cirrhosis has also been noted. In some studies the additional categories of chronic septal hepatitis and chronic lobular hepatitis have also been used. This approach has been useful but has a number of significant limitations. These include the difficulty of reliably recognising (or even defining) piecemeal necrosis coupled with the use of this single feature to distinguish the 2 main categories.

Furthermore other than noting the presence or absence of cirrhosis the extent of fibrosis is ignored. In the case of HCV these difficulties are rendered more significant because the piecemeal necrosis is often borderline and the lobular changes, which are ignored by this system, are more striking. For all these reasons this approach is passing out of favour.

More modern approaches to the histological classification of chronic hepatitis are based on the recognition that it is necessary to be able to fully describe both the activity of the inflammation and the progression of the fibrosis separately. They are termed grade and stage respectively (by analogy with reporting of large bowel cancer). This is logical because the severity of inflammation and fibrosis do not always parallel each other. One may get, e.g., active inflammation with little scarring or an inactive cirrhosis. The grade describes how active the liver disease is today while the stage describes how far down the road to cirrhosis the patient has gone. The concepts of grade and stage can be applied to other types of liver disease e.g. alcohol.

Studies of the natural history of chronic hepatitis and of new treatments have increasingly used the histological changes as an important endpoint. To facilitate their analyses semi-quantitative scoring systems have been developed. These have become increasingly used not only in large studies but also for the reporting of individual biopsies as they can provide a comprehensive picture of the histological changes.

The best known of these semi-quantitative scoring systems has been that of knodell et al. This scores portal inflammation, piecemeal necrosis (and bridging necrosis), lobular inflammation and fibrosis. Initially these 4 categories were added together to give a single figure, the Histological Activity Index (HAI). This has been subsequently modified in practice so that the 3 categories of inflammation are added together to give an overall measure of grade and the fibrosis (stage) reported separately. Despite this significant improvement a number of problems remain. These include a lack of precision in definitions of the different scores, the use of a discontinuous scoring system and only a limited attempt to assess reproducibility. Despite these criticisms the Knodell system has been widely used and has proved robust. The Scheuer scoring system was an attempt to deal with some of these criticisms but has not been widely used and has been overtaken by a modified version of the Knodell system called the modified HAI. This uses 5 categories (splitting bridging necrosis from piecemeal necrosis) and sensibly encourages the listing of histological aetiologies and in some cases may carry prognostic information. The recording of immunohistochemical findings is also suggested. This is mainly of importance with HBV where reagents are available for use on routinely fixed and processed material. Although this uses somewhat clearer definitions, a continuous scoring system and clearly separates stage and grade their is little experience with it and preliminary studies suggest inter-observer reproducibility is low. Another modification of the Knodell system has been developed by the METAVIR group. This study, importantly, confirmed the value of pathologists agreeing on the criteria to be used before embarking on a multicentre trial. It is worth noting that reproducibility remains an unresolved problem with all the scoring systems. Most studies suggest that, regardless of the system, there is better agreement for fibrosis than for inflammation.

HCV

HCV highlights a number of specific problems in addition to the ones described above. For example, hepatitis C produces more focal disease than HBV. This poses the question of how big a liver biopsy is adequate to reliably assess the severity of chronic hepatitis. As a rule of thumb a biopsy containing less than 3 portal tracts is inadequate but nevertheless when comparing biopsies this variability needs to be borne in mind. We have shown that in the cases of HCV the modified HAI correlates better with genotype than does the Knodell system. Given that a consensus is developing that genotype is an important predictor of both disease severity and response to treatment, this suggests that the modified HAI may be the more useful scoring system at least in the context of HCV. Fibrosis, like genotype, has been shown to be an independent predictor of response to interferon treatment. The question remains as to whether a single scoring system can be used for the diverse range of aetiologies. In the case of HCV specific histological features may carry prognostic information. For example in one study bile duct damage and the presence of stainable iron may correlate with response to treatment. Clinicians are playing increasing attention to the scores generated by pathologists. For example a protocol for managing patients with HCV based largely on the liver biopsy has been proposed and is being increasingly used. In this patients with significant fibrosis +/ inflammation are treated (those with cirrhosis are treated only if certain clinical criteria are met) while the remainder are followed up. The use of this type of protocol places additional responsibility on the pathologist.

Conclusion

It is clear that while progress has been made we do not yet have an ideal scoring system and much work remains to be done. This will require collaboration between groups of pathologists as well as close clinical collaboration with careful followup and the ability to not to adopt an entrenched position. The most important bench mark for any classification is that it should provide information predicting disease progression and the response to treatment. This has been conspicuous lacking in the past.

BIBLIOGRAPHY

  Hellenic Archives of Pathology, 1994, 8(3): 124-127