Immunohistochemistry and Special Stains in Liver Pathology

Immunohistochemistry and Special Stains in Liver Pathology

Michael S. Torbenson, MD


This chapter considers special stains used in medical liver specimens and is organized into the following sections: up-front stains, stains used to work up injury patterns, and miscellaneous stains used in the practice of medical liver pathology. Of course, stains are very important in tumor evaluation too, and they are covered in detail in their respective chapters.


There is little or no data on the stains that should be ordered up-front on every medical liver biopsy. In general, different pathologists use different up-front stains based on their training and the local practices. The up-front stains can range from a very limited panel of one hematoxylin and eosin stain (H&E) and a trichrome stain, to a comprehensive panel of multiple H&Es, trichrome, iron, periodic acid Schiff without diastase (PAS), periodic acid Schiff with diastase (PASD), copper, and reticulin stains. Which end is better? Perhaps neither is perfect as a routine up-front panel. One reasonable approach is shown in Table 4.1.

H&E stains are critical and should be ordered on every case. Empiric experience provides compelling evidence that two H&E stains are better than one. Almost all pathologists can recall cases where a critical lesion was only identified on one of the H&Es.

Table 4.1 Routine up-front stains in medical liver pathology


Use in routine practice

H&E × 2

Every case

Trichrome (or similar)

Every case


Baseline biopsy, others as needed


Baseline biopsy, others as needed


As needed based on H&E findings


As needed based on H&E findings

Trichrome stain overview

Trichrome stains, or similar connective tissue stains such as Sirous Red, are critical for evaluating fibrosis. In resource-limited environments, a reasonable alternative approach is to cut additional blanks when the H&E is cut so that a trichrome stain can be ordered if needed. Cutting additional blanks at the same time as the blank for the H&E is cut is important, because there can be substantial tissue lost each time the block is faced in histology. There are a few situations where a trichrome stain is not necessary. First, if the H&E clearly shows established cirrhosis, then the trichrome stain is unnecessary. Second, follow-up biopsies performed soon after an initial biopsy may not need a trichrome. This situation arises most clearly in the transplant setting, where some centers will perform a follow-up biopsy to evaluate for response to therapy within days to weeks after an initial biopsy that showed acute cellular rejection. In this setting, a trichrome stain on the follow-up biopsy is rarely needed.

Iron stain overview

Iron stains are critical tools to look for iron overload. The Perls iron stain is fairly sensitive and very specific for detecting hemosiderin. It is the most widely used stain for iron evaluation, and most centers perform this stain on every medical liver biopsy. The Perls iron stain can also highlight ferritin as a light blue cytoplasmic blush, but this can be ignored and attention is focused on the granular deposits of hemosiderin.

In resource-limited environments, it would not be unreasonable to examine the H&E carefully and then order an iron stain if there is visible pigment on the H&E that suggests iron, if the serum ferritin is significantly elevated (>1,000), transferrin saturation is 45% or greater, or there are other clinical reasons to be concerned about iron overload disease. This approach admittedly would miss almost all biopsies with minimal iron accumulation and many biopsies with mild iron accumulation. However, minimal to mild iron accumulation in most cases has no clinical significance.

For clinical purposes, the amount of iron can be reported either descriptively (minimal, mild, moderate, marked) or by using a numerical system (0, 1+, 2+, 3+, 4+). In 1962, Scheuer proposed one of the first grading systems for evaluating hepatic iron using the Perls iron stain. Many additional iron grading systems have been proposed over the years, all providing largely similar semiquantitative data. The different systems vary in their approach, with some systems scoring the iron stain by the zonation of hemosiderin distribution, some systems scoring by the lowest magnification that discernable granules of hemosiderin can be seen, and some by the percent of hepatocytes positive for hemosiderin. Is one system clearly the best for clinical care? Not really, as they all seem to work fine. After all, hemosiderin accumulation has the greatest clinical relevance when it is moderate or marked, levels that are captured well by all grading systems. A numerical system is important in research studies, so the authors can do statistical analysis and generate a figure for the paper. However, for routine clinical care, a numerical grade is not necessary and your report will be perfectly fine using descriptors (“mild,” “moderate,” etc.) and the iron location (hepatocellular vs. Kupffer cell). Nonetheless, there are many reasonable grading systems to choose from if you prefer.

Reticulin stain overview

In medical liver biopsies, reticulin stains are most helpful in identifying nodular regenerative hyperplasia (Fig. 4.1). Nodular regenerative hyperplasia represents a continuum of changes, easy to diagnose when well developed, and equally easy to over diagnose when not. If there isn’t clear nodularity on the H&E, make a diagnosis of nodular regenerative hyperplasia with caution. By definition, there should be no significant fibrosis, in particular bridging or worse. Some mild portal fibrosis can be seen in occasional cases. It’s also worth checking the liver enzymes, looking for a disproportional elevation in alkaline phosphatase levels, a pattern typically found with nodular regenerative hyperplasia.

Some pathologists also find reticulin stains helpful in separating bridging necrosis from bridging fibrosis, with bridging necrosis showing reticulin fibers, which result from the collapsed and compressed reticulin meshwork normally present in the lobules. In contrast, fibrosis will have less reticulin deposits. However, most pathologists use the H&E and the trichrome to make this distinction. In the workup of well-differentiated hepatocellular neoplasms, reticulin stains play a critical role in separating benign from malignant hepatocellular proliferations.

Periodic acid Schiff without diastase overview

A PAS stain without diastase highlights glycogen but is not very helpful for most diagnostic purposes. Some individuals find the stain helpful to identify foci of lobular necrosis, which will stand out against the PAS-positive hepatocytes (Fig. 4.2). Another use is in the workup for ground-glass-type changes in the hepatocytes, where a negative PAS stain suggests fibrinogen storage disease.

Figure 4.1 Reticulin stain. Nodular regenerative hyperplasia. The nodularity results from zones of atrophic hepatocytes juxtaposed with those that are normal in size and form ill-defined nodules.

Figure 4.2 PAS stain. Scattered foci of lobular inflammation stand out against the PAS stain.

Periodic acid Schiff with diastase overview

PASD stains are helpful in identifying the intrahepatic globules of α-1-antitrypsin deficiency (Fig. 4.3). The globules can be easily overlooked on H&E, but are nicely highlighted on a PASD stain. They typically have a periportal distribution in noncirrhotic livers, but can be patchy, so the entire slide should be examined. Although not necessary, PASD stains also nicely highlight macrophages in areas of lobular inflammation (Fig. 4.4). Finally, PASD stains can highlight small Kupffer cell globules composed of
immunoglobulin in the setting of autoimmune hepatitis (Fig. 4.5), primary biliary cirrhosis, or other diseases with hypergammaglobulinemia.

Figure 4.3 PASD stain. Numerous globules of α-1-antitrypsin are seen in this case; most biopsies don’t have so many globules.

Figure 4.4 PASD stain. Scattered foci of lobular macrophages are highlighted.

Figure 4.5 PASD stain. Small globules composed of immunoglobulins are seen in the Kupffer cell cytoplasm of this biopsy, which also showed autoimmune hepatitis.


Chronic cholestatic liver disease


Bile duct loss is best assessed using a keratin stain. As a stand-alone stain, CK AE 1/3 works well, but CK7 works just as well for bile duct loss and also highlights intermediate hepatocytes, a typical feature found with ductopenia. As a general rule of thumb, at least 50% of the portal tracts should lack bile ducts to confidently diagnose ductopenia. The alkaline phosphatase will be disproportionately elevated, a useful additional piece of information before making a diagnosis of ductopenia.

Chronic cholestasis

There are two primary stains used to support a diagnosis of chronic cholestatic liver disease: the rhodanine copper stain and the CK7 immunostain. Both are useful in noncirrhotic livers, but in cirrhotic livers these stains are not specific for biliary tract disease.

Copper is normally excreted in the bile. With chronic cholestasis, copper accumulates in the lysosomes of periportal hepatocytes and can be detected with a copper stain (Fig. 4.6). The staining can be very focal, so the stain has to be examined carefully. CK7 is the second stain used to support a diagnosis of chronic cholestasis. CK7 normally stains the bile ducts and bile ductules, but the hepatocytes will be negative. However, in the setting of chronic cholestatic liver disease, CK7 stains the periportal hepatocytes (Fig. 4.7). These positive staining hepatocytes are called intermediate hepatocytes. Often, CK7-positive intermediate hepatocytes are evident on liver biopsies before copper accumulation is found. However, specimens can be negative for CK7 but positive for copper in long-standing, low-grade chronic cholestasis. CK7 and copper stains are often best used together and should always be interpreted in conjunction with the H&E findings. Finally, in congestive hepatopathy, the CK7 positivity in hepatocytes can show a zone 3 distribution (Fig. 4.8).1

Inherited chronic cholestatic liver diseases include the progressive familial intrahepatic cholestasis diseases, types 1, 2, and 3. Type 2 is caused by mutation in the ABCB11 gene, which encodes the bile salt export pump (BSEP) protein. Immunostains for BSEP are not
widely available, but loss of the normal canalicular staining suggests ABCB11 mutations. However, loss of staining is not very sensitive or specific, so must be confirmed by sequencing studies.2, 3, 4 Likewise, PFIC3 is caused by mutations in the ABCB4 gene, which encodes the multiple drug resistance 3 protein. The MDR3 immunostain is also not very sensitive or specific, but loss of canalicular staining suggests ABCB4 mutations, which can then be confirmed by sequencing studies.

Figure 4.6 Rhodanine. Focal copper deposition is seen in the periportal hepatocytes in this case of primary sclerosing cholangitis.

Figure 4.7 CK7 immunostain. Periportal hepatocytes (intermediate hepatocytes) are positive in this case of chronic cholestatic liver disease.

Figure 4.8 CK7 immunostain. The zone 3 hepatocytes are positive in this case of congestive hepatopathy.

Autoimmune conditions of the liver

Autoimmune hepatitis

The plasma cells in autoimmune hepatitis are primarily immunoglobulin G (IgG)-positive, with much fewer numbers of immunoglobulin G (IgM)-positive plasm cells. In contrast, the portal tracts in primary biliary cirrhosis have equal or greater numbers of IgM-positive plasma cells, compared to IgG.5 In overlap syndromes, the IgG and IgM levels are generally similar, so immunophenotyping of the plasma cells is not helpful in separating primary biliary cirrhosis alone from an overlap syndrome. PASD stains can highlight small round globules in the Kupffer cells.

Primary biliary cirrhosis

The histological findings in early primary biliary cirrhosis can be very patchy, so additional H&E sections are often very helpful if the initial H&E findings are either almost normal or show only mild, nonspecific changes. Copper stains and CK7 stains are also very helpful in identifying evidence of chronic cholestasis and can be positive even when the H&E findings are almost normal.

As noted above, the portal tracts in primary biliary cirrhosis have equal or greater numbers of IgM-positive plasma cells, compared to IgG. This finding correlates very well with serum findings, where primary biliary cirrhosis is also characterized by elevated serum IgM and IgG levels, whereas autoimmune hepatitis is characterized by elevated serum IgG levels. Immunostains for CD1a (Figs. 4.9 and 4.10) can also be helpful in some cases of primary biliary cirrhosis, because dendritic cells in the inflamed bile ducts can be positive.6 PASD stains can also highlight small round globules in the Kupffer cells (See Fig. 4.5).

Only gold members can continue reading. Log In or Register to continue

Nov 24, 2019 | Posted by in GASTROENTEROLOGY | Comments Off on Immunohistochemistry and Special Stains in Liver Pathology
Premium Wordpress Themes by UFO Themes