Approach to the Liver Biopsy in the Patient With Chronic Low-Level Aminotransferase Elevations

There is extensive literature addressing both the histopathologic and clinical characteristics of every liver disease entity that presents with specific clinical and/or laboratory findings. Many patients, however, are biopsied not because of suspicion for a specific disease, but because they have low-level aminotransferase elevations without any clinical signs or symptoms.

A few of these biopsies are entirely normal, and the pathologist can confidently diagnose these biopsies as such. However, many of these biopsies are not completely normal but have only minor, nonspecific histologic changes. For example, these biopsies sometimes contain minute nonnecrotizing granulomas (Figure 1, A). Low-grade inflammation is often present in a few portal tracts, possibly with a few plasma cells or eosinophils (Figure 1, B), whereas other portal tracts may be entirely normal. Rare necrotic hepatocytes (Figure 1, C), debris-laden macrophages (Figure 1, D), and small foci of inflammation (Figure 2, A) may be seen in the lobular parenchyma. Sometimes there is focal steatosis (Figure 2, B). Hepatocytes may be enlarged and pale, and there may be variation in size of the hepatocytes and their nuclei with occasional binucleate cells (Figure 2, C). There may be mild pericentral sinusoidal dilatation (Figure 2, D).

Most of the literature addresses specific disease entities, and textbooks are commonly organized around categories of disease, with maybe some introductory chapters discussing what is normal. This leaves only limited literature or textbook guidance on how best to handle cases with some minor abnormality or abnormalities similar to those listed above, but without major changes that would help the pathologist to make a specific diagnosis.

Before addressing how to make a diagnosis in these cases, it is important to discuss some common mistakes pathologists make in signing out these biopsies. Lacking a specific disease entity to assign a case to, but noticing that it is not entirely normal, the pathologist may diagnose the case as “mild chronic hepatitis” or something similar. Unfortunately, making such a diagnosis is not helpful to the clinician; “mild chronic hepatitis” has no signs, symptoms, or specific pathologic or laboratory findings, as opposed to actual diseases, which have one or all of these. As pathologists, we must remember our primary job is to be consultants and use our knowledge and expertise to guide clinical care, and a diagnosis of mild chronic hepatitis does not provide useful information. In addition, the use of the word hepatitis may falsely tell the clinician that there is a disease, and the patient will be needlessly worried or even subjected to unnecessary additional studies and/or treatment as a result.

Another temptation when signing out this biopsy is to extrapolate minor findings to raise possibilities for which there is no real evidence. For example, a pathologist might see a positive antinuclear antibody and 3 plasma cells in a portal tract somewhere in the liver biopsy from one of these asymptomatic patients with elevated transaminases, and reason that because autoimmune hepatitis also has plasma cells and autoantibodies with elevated aminotransferases, the possibility of autoimmune hepatitis should be raised in a comment. Although it may seem cautious and prudent to list every remote possibility on first glance, this temptation should also be resisted; it again subjects the patient to expensive and unnecessary treatment and further testing based on essentially no evidence.

When the judicious pathologist encounters a difficult biopsy and wants to determine how to best help the clinician in treating the patient, a logical next step is to consult the literature for information that could be helpful in assisting with the diagnosis. Perhaps these changes indicate a known disease, or at least lead to a differential diagnosis that includes a few diseases. Unfortunately, as hinted at earlier, even with decades of literature and numerous manuscripts written every year on hepatology or liver pathology, there is almost no literature on either the pathology or the clinical features of the nearly normal liver biopsy in the asymptomatic patient with chronic mildly elevated aminotransferases.

The earliest mention in the literature regarding the clinical and pathologic features of a cohort including a substantial portion of near-normal liver biopsies is a 1965 study by Steigmann et al.¹  This study followed 103 patients diagnosed with “nonspecific (reactive) hepatitis” on biopsy regardless of the indication for the biopsy. Nonspecific (reactive) hepatitis was defined as a nonspecific inflammatory process with focal necrosis, granulocytic or lymphocytic inflammation, Kupffer cell proliferation, and at least some acute or chronic inflammatory cells in the portal tract. These findings are each commonly seen in biopsies from asymptomatic patients and low-level liver enzyme elevation, indicating a substantial overlap between biopsies diagnosed as “nonspecific (reactive) hepatitis” and our group of nearly normal liver biopsies. The authors cited a 1957 definition of nonspecific (reactive) hepatitis by Hans Popper and Fenton Schaffner²  in their textbook Liver: Structure and Function; their description appears to be the earliest in either textbooks or medical literature attempting to group together the disparate minor histologic findings in patients without a known clinical liver disease.

Clinically, many of these patients were not tested for alanine aminotransferase or aspartate aminotransferase levels, and most of those who were tested had normal or near-normal aminotransferase levels. Only one-third of these patients were admitted with a diagnosis of some type of liver disease; the remaining patients were admitted with other diseases, including pulmonary, metabolic, and gastrointestinal tract diseases, and approximately 15% of patients were biopsied out of concern for liver metastasis. On a “short” follow-up, no hepatic sequelae had been identified among any of the patients in the study of Steigmann et al,¹  and unfortunately a literature search reveals no long-term follow-up. This cohort may have some level of overlap with our group of patients with low-level aminotransferase elevations and without any liver-related symptoms. However, given that patients were selected for this study based on biopsy findings without consideration of clinical scenario, it is impossible to tell how many patients, if any, would have chronic low-level aminotransferase elevations if these laboratory values were routinely followed. Of note, hepatitis C virus would not be discovered until 1989; given that it is often asymptomatic with similar histologic findings in its early stages, it likely accounted for at least some of these patients. These factors limit the usefulness of this study to the modern-day pathologist who wishes to guide the clinician to the correct management and follow-up.

A publication from the Mayo Clinic by Hay et al³  in 1989, still just before the discovery of the hepatitis C virus, specifically focused on patients with unexplained chronic aminotransferase elevations and no symptoms, but the biopsy findings were able to provide specific diagnoses in most of these patients. Of 47 patients, 34 had chronic hepatitis; of these, 16 were cirrhotic and 18 had elevated anti-nuclear antibodies or anti–smooth muscle antibodies; the manuscript does not specify which of these biopsies had autoimmune hepatitis. Presumably some of these chronic hepatitis patients also had hepatitis C infection. Of the patients without chronic hepatitis, 10 had steatohepatitis and 3 had miscellaneous disorders.

Ten years later, in 1999, now with the knowledge of hepatitis C virus in hand, Mathiesen et al⁴  published the next study of asymptomatic patients with unexplained aminotransferase elevations. This study followed a cohort of 150 patients, of whom 5 had what was described as normal histology with “minor unspecific changes.” Of the remaining patients, 23 were diagnosed with chronic hepatitis C; 75 had steatosis, steatohepatitis, or some other presumed alcoholic liver disease; 36 had chronic hepatitis of unknown etiology, of which cases 22 were considered mild.

In order to best guide clinician management on biopsies from patients with elevated aminotransferase levels and no symptoms, it is helpful to understand the decision-making process that led to the liver biopsy. There is a flowchart in chapter 73 of the 10th edition of Sleisenger and Fordtran's Gastrointestinal and Liver Disease⁵  for evaluation of this group of patients. The first step is a review of the patient's history, including medication list and alcohol and illegal drug use. If the history does not reveal a cause, the next step is to perform tests for hepatitis viral serology, iron studies, and tests for autoimmune markers and other markers of liver disease. Liver ultrasound is also useful in order to rule out fatty liver disease. If these tests still do not provide an answer, finally liver biopsy is performed for only those patients who continue to have aminotransferase levels at least twice the upper limit of normal after an otherwise negative workup.

In a selection of recent liver pathology textbooks, most books at least briefly address liver biopsies with only minor, nonspecific changes. However, it takes some effort to find this information in these books, and attention is devoted only briefly to these common biopsies. The most recent 7th edition of MacSween's Pathology of the Liver^{6(p966–967)} contains a summary of “nonspecific reactive hepatitis,” the same entity described by Popper and Schaffner²  in their textbook more than 60 years ago. The description is largely the same as the initial description from decades ago, although there is a note that hepatitis C virus should be excluded. Histologic features are largely the same as those seen in liver biopsies from asymptomatic patients with elevated aminotransferases, including focal necrosis, varying size and shape of liver cells, and variable portal chronic inflammation, among other changes.

However, although the histologic description is similar, no clinical implications of these minor changes are listed except that they are nonspecific. Similarly to the 1965 study of Steigmann et al,¹  the biopsies include a disparate group of clinical scenarios, including but not limited to asymptomatic patients with aminotransferase elevation.¹  Again, this entity is seen in many cases with an entirely different biopsy indication, such as resolving hepatitis, changes near the lesion in attempted biopsies of liver masses, recent febrile illness, and inflammation elsewhere in the splanchnic bed. Overall, this chapter leaves us with a helpful description of a histologic entity but without any clinical implications.

Suriawinata and Thung's Liver Pathology: An Atlas and Concise Guide takes a similar but not entirely identical approach.⁷  Nonspecific reactive hepatitis again includes a group of similar findings to the minor nonspecific changes often seen in biopsies from asymptomatic patients with elevated aminotransferases. However, the entity is separated from cases of residual hepatitis with prominent clusters of debris-laden macrophages indicating recent hepatic necrosis. Clinically, the authors associate nonspecific reactive hepatitis with systemic disease, stating that prognosis and treatment in biopsies with these findings will depend on features of an underlying systemic disease. Hence this definition applies to biopsies that may have similar histologic findings to our biopsy in question, but within a different clinical context where presumably there is some systemic disease and not just an elevated aminotransferase level of unknown significance.

In addition to the difference in clinical context between the above definitions of nonspecific reactive hepatitis and the biopsy with minor histologic changes in an asymptomatic patient with elevated aminotransferases, the latter group of biopsies does not always have inflammation, making a diagnosis of hepatitis inappropriate.

Torbenson's⁸  third edition of Biopsy Interpretation of the Liver devotes a full chapter to the “almost-normal liver biopsy.” The chapter includes biopsies with subtle but still diagnostic abnormalities and a brief description of cryptogenic cirrhosis, which can also be identified on biopsies of patients with unexplained aminotransferase elevations. More important to the question at hand, there is a brief discussion in the final 2 pages of the chapter addressing mostly near-normal liver biopsies from asymptomatic patients with elevated liver enzymes, although it also includes patients with unexplained ascites. Several possible clinical associations are listed and will apply to about half of these patients, and liver enzymes will eventually normalize in another fraction.

Lefkowitch⁹  devotes a single paragraph to these biopsies in his 9th edition of Scheuer's Liver Biopsy Interpretation under the heading of “non-specific reactive changes.” The paragraph mentions changes including Kupffer cell hypertrophy, focal hepatocellular necrosis, steatosis, and portal and lobular inflammation. The author notes that clinical information can help distinguish cases of residual acute hepatitis or mild chronic hepatitis from those biopsies with nonspecific reactive changes in systemic diseases. Overall this clinical and histologic summary recalls the diagnosis of nonspecific reactive hepatitis described in other texts, with the exception that the word hepatitis is not used because not all of these cases have the inflammation required for a diagnosis of hepatitis.

In Liver Pathology from the Consultant Pathology series, chapter 19 by Jain and Kakar¹⁰  goes over the approach to liver biopsies with minimal or nonspecific histologic findings. This is a useful chapter for the pathologist facing such a biopsy, but the focus is mostly on diseases with subtle but diagnostic changes and systemic diseases with minor histologic findings in the liver. In addition, there is a short section on nonspecific reactive hepatitis, which is described as a response to systemic illnesses. There is a brief but quite accurate comment that these cases are often frustrating, and a single paragraph addresses the fact that biopsy sometimes cannot explain the etiology of liver enzyme abnormalities.

One interesting finding in these various textbook chapters and in much of the literature addressing these biopsies is the comment that similar biopsies are often seen in patients with systemic diseases. This is certainly true, and as biopsy was once a common part of the workup for fever of unknown origin, we used to see these cases more commonly. However, although there may be similar minor abnormalities on biopsies from patients with fever of unknown origin and/or clear evidence of a systemic disease, the clinical scenario, management, and follow-up are vastly different compared with patients without symptoms who are biopsied only because of elevated aminotransferase levels. The textbooks reviewed above offer no references except Popper and Schaffner's²  definition of nonspecific reactive hepatitis in 1957, when many of these biopsies were presumably for fever of unknown origin whereas others were intended to biopsy a mass lesion.

Finally, in September 2017 Czeczok et al¹¹  published a study focusing on clinical associations and follow-up of normal and almost-normal liver biopsies in patients biopsied for either abnormal liver enzymes or unexplained ascites. Of their 97-patient cohort, only 77 (79%) had elevated liver enzymes, which was not quite identical to our scenario where the biopsy is for elevated transaminases in an asymptomatic patient. However, there is enough overlap for the results of the study to be informative. Seventy percent of the biopsied patients were found to have a clinical association with a systemic disease, including autoimmune disease in 18%, vascular disease or ischemic events in 13%, metabolic syndrome in 11%, drug effects in 8%, and gastrointestinal inflammatory diseases in 7% of patients. Twenty-eight percent of the patients were not found to have any clinical association, and in about half of these patients the liver enzymes normalized on follow-up, whereas the other half had persistent elevations in liver enzymes without any evidence of disease. Only 7% of patients in the study developed chronic liver disease.

In summary, although the past 6 ½ decades have produced limited textbook and literature information on the nearly normal liver biopsy in an asymptomatic patient with elevated transaminase levels, this information all points to the conclusion that this biopsy is not associated with any liver disease. It may be seen in patients with systemic disease, but these diseases have other findings and the liver is not the cause of symptoms in these cases. In other cases, there is an elevation in liver enzymes without any clinical or pathologic correlation, which may or may not normalize later. In all of these situations, the importance of this biopsy is that it is not diagnostic of any liver disease, nor does it lead to a differential diagnosis between multiple liver diseases. The pathologist faced with this biopsy can provide a simple diagnosis with or without a comment, stating the lack of any significant abnormalities and noting that there are only minor, nondiagnostic changes. This informs the clinician that the patient does not need to be worked up or followed up for any liver disease and provides the patient with peace of mind that he or she does not have a liver disease.