Application of Immunohistochemistry to Liver and Gastrointestinal Neoplasms

Liver, Stomach, Colon, and Pancreas

Stephen A. Geller, MD; Deepti Dhall, MD; Randa Alsabeh, MD

Context.Immunohistochemistry has become an integral component of the practice of pathology. Newer antibodies allow for increasingly precise diagnoses for tumors that previously could not be easily identied. Recently, immunohistochemical evaluations have begun to allow pathologists to actively assist in determining prognosis and even in selecting therapies. Objective.To summarize the usefulness of currently available immunostains for the study of liver and gastrointestinal system neoplasms and to make recommendations for panels of immunostains that can be particularly helpful. Data Sources.Information has been collected from re-

cent literature as well as from personal experience and practice. Conclusions.Many immunostains are now available for the practicing pathologist that allow for increasing accuracy in diagnosis of liver and gastrointestinal tract neoplasms. Panels of immunostains can be used to differentiate between various tumors and also to identify site of origin in the case of a metastatic neoplasm. Immunostains that allow for prognostic determinations and for guidance in the selection of chemotherapeutic agents can also be used by pathologists to assist in the management of patients with malignant tumors affecting the liver and gastrointestinal tract. (Arch Pathol Lab Med. 2008;132:490499) langiocarcinoma [CCA]), (2) distinguish primary hepatic tumors from metastases, and (3) determine site of origin of metastases. Hepatocellular Carcinoma Hepatocellular carcinoma can react with antibodies directed against cytokeratins (CKs), particularly low-molecular-weight CKs (CK8, CK18, and CAM 5.2), as well as with -fetoprotein, 1-antitrypsin, 1-antichymotrypsin, CD10, villin, and human hepatocyte parafn 1 (HepPar1).3 Hepatocellular carcinomas are usually negative with CK7, CK20, and AE1/AE3. These immunostains are not useful for the differentiation of benign from malignant hepatocytic lesions and are not specic for HCC. More specic is the demonstration of intercellular canaliculi with antibody directed against polyclonal carcinoembryonic antigen (pCEA) (Figure 1) because of cross-reactivity of pCEA with biliary glycoprotein 1. As many as 90% of HCCs will demonstrate canalicular pattern with pCEA.3 This is not seen with monoclonal carcinoembryonic antigen (mCEA) antibody. Demonstration of albumin is also specic for liver cells, but a useful antibody is not yet commercially available. Neither pCEA nor albumin immunoreactivity is absolutely pathognomonic for primary HCC because malignant tumors with hepatocellular differentiation arising in various extrahepatic sites (eg, stomach, uterus), so-called hepatoid carcinomas, can also demonstrate canaliculi with pCEA and albumin production.4 Further, HepPar-1 can be positive in many tumors in which the tumors cells have an eosinophilic, granular cytoplasm. Increasingly, immunohistochemical and molecular studImmunohistochemistry of Liver and GI NeoplasmsGeller et al

his review concentrates on the immunohistochemical phenotypes of the major neoplasms of the liver and gastrointestinal tract and also includes discussion of immunohistochemical and molecular markers that can be useful for therapy selection and prognostication. We do not emphasize lymphoproliferative disorders, vasoproliferative disorders, vascular tumors, rare tumors, and tumors unique to children. The hematoxylin-eosinstained slide remains the keystone of all diagnostic studies, with immunohistochemical reactions and molecular studies supplementing, but not supplanting, histopathology evaluations. It is also important to consider that experience in directing the performance of immunostains is vital in ensuring that interpretations are not based on technical artifacts. Not all commercial preparations of a given antibody will react in the same way. Similarly, expertise is required in the diagnostic evaluation of immunostains.1,2 LIVER Immunohistochemical reactions are principally applied to liver neoplasms to (1) differentiate between primary hepatic tumors (hepatocellular carcinoma [HCC] and choAccepted for publication July 5, 2007. From the Department of Pathology and Laboratory Medicine, CedarsSinai Medical Center, Los Angeles, Calif. The authors have no relevant nancial interest in the products or companies described in this article. Reprints: Stephen A. Geller, MD, Department of Pathology and Laboratory Medicine, Cedars-Sinai Medical Center, 8700 Beverly Blvd, Room 8728, Los Angeles, CA 90048-0750 (e-mail: stephen.geller@ cshs.org). 490 Arch Pathol Lab MedVol 132, March 2008

Figure 1. Hepatocellular carcinoma, with clear cell differentiation. A, Hematoxylin-eosin (original magnication noembryonic antigen showing canalicular pattern ( 400). Figure 2. Angiomyolipoma of the liver. A, Hematoxylin-eosin (original magnication

200). B, Polyclonal carci200). 40). B,

200). B, HMB-45 (original magnication

Figure 3. Glypican-3 immunostaining of hepatocellular carcinoma in a cirrhotic liver. A, Hematoxylin-eosin (original magnication Glypican-3 (original magnication 40). Arch Pathol Lab MedVol 132, March 2008

Immunohistochemistry of Liver and GI NeoplasmsGeller et al

491

Table 1.
HepPar-1 pCEA

Immunophenotype of Some Clear Cell Tumors That Can Be Found in the Liver*
CD34 RCCa Vimentin HMB-45 Melan-A CK5 CK8 Chromo Synapto TTF-1 S100 EMA CD117

HCC ( ) RCC ( ) ( ) ACC ( ) ( ) Mel ( ) CCSa ( ) ( ) Pulm ( ) ( ) ( ) ( ) SqCCa ( ) ( ) ( ) Neuren ( ) ( ) ( ) EpLeiS ( ) ( ) AML ( ) ( ) ( ) ( ) GIST ? ? ? ? ? ? ( ) ( ) ? ( ) * HepPar-1 indicates hepatocyte parafn 1; pCEA, polyclonal carcinoembryonic antigen; RCCa, renal cell carcinoma antibody; CK, cytokeratin; Chromo, chromogranin; Synapto, synaptophysin; TTF-1, thyroid transcription factor 1; EMA, epithelial membrane antigen; CD117, c-Kit; HCC, hepatocellular carcinoma; , more than 50% of cases; , negative; ( ), fewer than 50% of cases; RCC, renal cell carcinoma; ACC, adrenocortical carcinoma; Mel, melanoma; CCSa, clear cell sarcoma; Pulm, pulmonary carcinoma; SqCCa, squamous cell carcinoma; Neuren, neuroendocrine carcinoma; EpLeiS, epithelioid leiomyosarcoma; AML, angiomyolipoma; GIST, gastrointestinal stromal tumor; and ?, not studied.

ies are being used to augment more traditional approaches to predicting prognosis. Commonly available antibodies such as p53 and Ki-67 have been shown to be useful in this regard.5,6 As another example, overexpression of platelet-derived growth factor may indicate a high potential for metastasis in HCC.7 Fibrolamellar HCC Fibrolamellar HCC is immunophenotypically similar to the usual HCC but unlike HCC also stains strongly with CK7.8 1-Antitrypsin is more likely to be demonstrable in brolamellar HCC than in usual HCC.9 Rarely, brolamellar HCC can show neuroendocrine differentiation.10 Strong expression of epidermal growth factor receptor (EGFR) in brolamellar HCC may be predictive of response to treatment with EGFR antagonists.11 Clear Cell HCC Clear cell HCC can be histologically indistinguishable from adrenal cortical carcinoma, renal cell carcinoma, and neuroendocrine carcinoma. Adrenal cortical carcinoma often stains with inhibin and Melan-A and is generally nonreactive to HepPar-1.12 Adrenal cortical carcinoma may also show minimal staining with CAM 5.2. Both adrenal cortical carcinoma and renal cell carcinoma can be composed mostly of clear cells, as can HCC (Figure 2). Renal cell carcinoma can be identied by reactivity to vimentin as well as to antibody prepared against renal cell carcinoma cells. In HCC, vimentin is generally seen only in high-grade HCC with spindle cell differentiation. A variety of other clear cell tumors can rarely involve the liver, and immunostains can be helpful in their differentiation (Table 1). Neuroendocrine carcinomas can sometimes be predominantly clear cell but most often have a trabecular and acinar pattern resembling the usual HCC and demonstrate reactivity with neuroendocrine-associated antibodies, including chromogranin and synaptophysin. Cholangiocarcinoma Cholangiocarcinoma most often must be differentiated from metastatic carcinoma. Cholangiocarcinoma is almost always a desmoplastic tumor and often is mucin producing. The immunophenotype of CCA is generally similar to that of extrahepatic pancreaticobiliary carcinomas. Almost all CCAs react with CK7, CK19, CEA (both monoclonal and polyclonal; noncanalicular pattern), and MOC-31. In
492 Arch Pathol Lab MedVol 132, March 2008

contrast to other pancreaticobiliary carcinomas, CCAs are more likely to react with CK7 and less likely to react with CK17, CK20, and p53. Cytokeratins can also be useful in differentiating HCC from CCA, as can thyroid transcription factor 1 and claudins.1315 Cholangiocarcinoma can also react with CA 125, and differentiation from mullerian carcinomas can be difcult, although CCA is almost never estrogen receptor positive.16 Newer studies with the mucins (eg, MUC4, MUC5AC, MUC5B, MUC6) suggest that they may be useful in the subclassication of CCAs17 and may be useful in evaluating prognosis.18,19 Combined HCC and CCA Uncommonly, primary liver tumors show both HCC and CCA differentiation. In our experience, the CCA component is usually more aggressive and recurrences will often reect only CCA differentiation. We apply this diagnosis when there is both hematoxylin-eosin patterns of HCC and CCA and when the immunohistochemical expressions of these distinctive areas are similarly distinctive, with usual HCC immunostaining in the HCC areas and usual CCA expression of the CCA pattern. Whether or not the CCA originates from the HCC or the HCC from the CCA, as forms of malignant metaplasia, remains unresolved.20,21 Epithelioid Hemangioendothelioma The neoplastic cells of epithelioid hemangioendothelioma tend to vary greatly, with epithelioid, vacuolated, and dendritic forms. They typically stain with CD31, CD34, and factor VIII. Fewer than 50% will react with AE1/AE3, CAM 5.2, S100, or smooth muscle actin. Angiomyolipoma Angiomyolipoma can arise in the liver or can metastasize from the kidney. When composed principally of smooth muscle cells, angiomyolipoma can strongly resemble both HCC and renal cell carcinoma (Figure 3). The smooth muscle cells usually have weak staining with muscle markers such as actin and desmin and patchy strong staining with melanoma markers, including HMB-45 and Melan-A22,23 (Figure 2). The cells are typically CK negative and vimentin positive, in contrast to the usual HCC. Most angiomyolipomas will also have estrogen receptor positivity.24
Immunohistochemistry of Liver and GI NeoplasmsGeller et al

Table 2.

Unknown Primary Screening PanelAdult Males*

CK7 Lung, pancreaticobiliary, thyroid, mesothelioma CK20 Colon, stomach TTF-1 Lung, thyroid Renal cell carcinoma Renal cell carcinoma Villin Colon, stomach, pancreaticobiliary, hepatocellular carcinoma HepPar-1 Hepatocellular carcinoma CDX-2 Colon, ovarian S100 Melanoma, neuroendocrine, lung HMB-45 Melanoma, angiomyolipoma Melan-A Melanoma, angiomyolipoma Prostate-specic antigen (PSA) Prostate Prostatic alkaline phosphatase (PAP) Prostate CD45, CD3, and CD20 Lymphomas * CK indicates cytokeratin; TTF-1, thyroid transcription factor 1; HepPar-1, hepatocyte parafn 1. Metastases from prostate to liver are rare in the absence of metastasis to other sites (eg, bone).

Table 3.

Unknown Primary Screening PanelAdult Females*

CK7 Lung, pancreaticobiliary, thyroid, mesothelioma CK20 Colon, stomach TTF-1 Lung, thyroid Renal cell carcinoma Renal cell carcinoma Villin Colon, stomach, pancreaticobiliary, hepatocellular carcinoma HepPar-1 Hepatocellular carcinoma CDX-2 Colon, ovarian S100 Melanoma, neuroendocrine, lung HMB-45 Melanoma, angiomyolipoma Melan-A Melanoma, angiomyolipoma Estrogen receptor Breast, gynecologic GCDFP-15 Breast Bcl-2 Breast, gynecologic, neuroendocrine, melanoma, thyroid CD45, CD3, and CD20 Lymphoma * CK indicates cytokeratin; TTF-1, thyroid transcription factor 1; HepPar-1, hepatocyte parafn 1; and GCDFP-15, gross cystic disease uid protein 15.

Table 4.

Common Immunohistochemical Patterns in Differentiating Hepatocellular Carcinoma From Metastatic Adenocarcinoma*

Hepatocellular Carcinoma Metastatic Adenocarcinoma

HepPar-1 CEA, monoclonal / CEA, polyclonal (canalicular pattern) (noncanalicular pattern) CK7 CK20 CD10 (canalicular pattern) (noncanalicular pattern) Villin (canalicular pattern) (noncanalicular pattern) MOC-31 -Fetoprotein / CD34 (sinusoidal pattern) / (vascular pattern) * HepPar-1 indicates hepatocyte parafn 1; , strongly positive ( 50% of cells); , negative; CEA, carcinoembryonic antigen; / , weakly positive ( 10% of cells); , positive (10%50% of cells); and CK, cytokeratin. Modied from Lau et al. 3 Lung, pancreaticobiliary, thyroid, mesothelioma. Colon, stomach.

Adenoma and Focal Nodular Hyperplasia Differentiation between the relatively rare hepatic cell adenoma and focal nodular hyperplasia can be difcult in biopsy material.25,26 -Catenin will be expressed in hepatic cell adenoma.27,28 Differentiation between adenoma and HCC can also be problematic, although it is likely that increasingly available genetic studies will help resolve this problem.29 Immunostaining for glypican-3 may also prove to be helpful in this regard, with HCC expressing significantly more glypican-3 than cirrhotic nodules or benign liver cell tumors (Figure 3).30 Metastatic Carcinoma The most common use of immunohistochemistry in the study of liver tumors is to identify the site of origin of a
Arch Pathol Lab MedVol 132, March 2008

metastatic tumor when the primary site has not yet been identied (Tables 2 and 3). The development and implementation of a panel of immunostains can help resolve almost all diagnostic problems.3,3134 Cytokeratin 7 and CK20 are the rst step in the immunohistochemical identication of many tumors, with additional immunostains, some relatively specic for tumors of males and of females, to identify potential sites of origin (Tables 4 and 5). Panels can be also be developed to evaluate specic tumor patterns, as in the example of clear cell tumors discussed previously. STOMACH Immunohistochemical studies are generally not needed for the evaluation of benign and malignant epithelial tuImmunohistochemistry of Liver and GI NeoplasmsGeller et al 493

Table 5. Immunohistochemical Findings in Hepatocellular Carcinoma (HCC), Cholangiocarcinoma (CCA), and Metastatic Adenocarcinoma*
HCC, % CCA, % Metastatic ACA, %

Gastrointestinal Stromal Tumors CD117 stains most cases of gastrointestinal stromal tumor, including metastases.4851 Although there may sometimes be variation in distribution of CD117 positivity within a given tumor, in most cases staining is diffuse (Figure 4). When CD117 is positive in tumors other than gastrointestinal stromal tumor, the staining is almost always patchy. CD34 staining can also be seen in gastrointestinal stromal tumor. COLON Adenomas and Adenocarcinoma Adenomas (tubular adenoma, tubulovillous adenoma, villous adenoma) demonstrate the same immunohistochemical reactions as colonic adenocarcinoma. Almost all react with antibody directed against CK20 and a minority will also stain focally with CK7, in contrast to pancreatic adenocarcinomas, most of which are CK20 negative and CK7 positive. Differentiation of metastatic colorectal adenocarcinoma from adenocarcinoma arising at other sites can sometimes be challenging. Pulmonary adenocarcinoma can resemble colorectal adenocarcinoma. Cytokeratin 7 and CK20 can be helpful in this regard, with CK7 usually strongly positive in lung lesions and CK20 usually negative; the reverse pattern is seen with colorectal adenocarcinoma (Table 6). In addition, thyroid transcription factor 1 is generally positive in lung cancers, and CDX-2 and -catenin are generally positive in colorectal cancers. Endometrioid-type carcinomas can also be histologically indistinguishable from colorectal carcinoma. Here, again, CK7 is positive in almost all endometrioid adenocarcinomas and only mildly reactive in colorectal adenocarcinomas (Table 6). Cytokeratin 20 is generally negative for lung primaries but positive for colorectal. Appendiceal Adenocarcinoma Appendiceal adenocarcinoma will typically show staining for MUC5AC, in contrast to colonic adenocarcinoma in which this antibody is rarely reactive.52,53 This is particularly useful in studying mucinous adenocarcinomas that have metastasized in the abdomen. -Catenin is another differentiating antibody, positive in almost all colonic adenocarcinomas and negative in appendiceal adenocarcinomas. In women with abdominal mucinous carcinomatosis, distinction of colonic and appendiceal adenocarcinoma from ovarian adenocarcinoma is important. In colonic tumors, both villin and -catenin are often expressed; in appendiceal metastases villin is often expressed, but -catenin is unusual and in ovarian mucinous adenocarcinomas neither villin nor beta catenin are seen.54 Similar to appendiceal lesions, ovarian carcinomas express MUC5AC and similar to colorectal adenocarcinoma, ovarian mucinous adenocarcinomas express CDX-2. Useful supplements to the basic panels for unknown primaries in which appendiceal or ovarian mucinous tumors are suspected are MUC5AC and -catenin. Villin can also be helpful because it typically has a brush-border pattern of staining in both colonic and appendiceal adenocarcinoma and is typically cytoplasmic in ovarian and pancreatic lesions.55
Immunohistochemistry of Liver and GI NeoplasmsGeller et al

HepPar-1 90 0 AE1/AE3 45 89 CAM 5.2 98 89 CK7 21 78 CK19 10 44 CK20 5 11 * ACA indicates adenocarcinoma; HepPar-1, hepatocyte and CK, cytokeratin. Modied from Lau et al.3

14 95 89 36 29 30 parafn 1;

mors of the stomach because the histopathology is generally diagnostic but are used in the study of metastatic gastric carcinoma when the site of origin is not clear or when the macroscopic/radiologic appearance of the tumor is confusing (eg, gastric carcinoma directly and massively invading the liver and histologically indistinguishable from CCA). Adenocarcinoma Gastric adenocarcinomas will react with many antibodies directed against keratins, including AE1/AE3, CAM 5.2, CK18, CK19, CK7, and CK20. When CK7 and CK20 are used together, many gastric adenocarcinomas will stain more with both CK7 and CK20.33,3537 Approximately 25% will be positive for one and negative for the other (eg, CK7 /CK20 , CK7 /CK20 ), and a small number of cases will be negative for both. CDX-2, initially thought to be specic for colon carcinoma, will be reactive in more than 50% of cases38,39 and may be indicative of a lesser degree of invasiveness.40,41 Even HepPar-1, a useful marker for hepatocytes, will be positive in more than 50% of gastric cancers, including signet-ring cell carcinoma.42 The quantity and quality of mucus production by gastric carcinoma, as evaluated by immunohistochemical study of mucins, may be prognostically important; MUC2 expression is associated with poor survival.43 Neuroendocrine Carcinoma Neuroendocrine carcinoma characteristically stains with synaptophysin, chromogranin, villin, and CD57.44,45 In contrast to gastric adenocarcinoma, carcinomas occurring in the second part of the duodenum may be negative for both synaptophysin and chromogranin but will often react with somatostatin. The proliferation marker Ki-67 and the adhesion molecule E-cadherin have been used to assess aggressiveness of neuroendocrine carcinoma.46 A high ( 60%) Ki-67 proliferation index predicts aggressive behavior, and loss of E-cadherin may predict lymph node metastasis. Gastrointestinal Adenocarcinoma With Neuroendocrine Differentiation Gastric adenocarcinomas, both intestinal type and signet-ring cell type, can have neuroendocrine differentiation that may not be obvious with hematoxylin-eosin staining but will show staining with chromogranin and synaptophysin.47
494 Arch Pathol Lab MedVol 132, March 2008

Figure 4. Gastrointestinal stroma tumor. A, Hematoxylin-eosin (original magnication magnication 200).

200). B, CD117 (c-Kit) showing diffuse staining (original 100). B, Epidermal growth factor 200). B, Chymo495

Figure 5. Colonic adenocarcinoma metastatic to the liver. A, Hematoxylin-eosin (original magnication receptor showing membranous staining (original magnication 100).

Figure 6. Acinar cell carcinoma of the pancreas with neuroendocrine features. A, Hematoxylin-eosin (original magnication trypsin (original magnication 400). Arch Pathol Lab MedVol 132, March 2008

Immunohistochemistry of Liver and GI NeoplasmsGeller et al

Table 6. Differentiation of Colorectal Adenocarcinoma From Pulmonary and Endometrioid Adenocarcinomas*

Colorectal Pulmonary Endometrioid

for the presence of MSI with abnormal immunohistochemistry results is 100%.61,62 PANCREAS Invasive Ductal Adenocarcinoma Adenocarcinoma of the pancreas usually derives from precursor stages of pancreatic ductal dysplasia. The immunostaining pattern of high-grade pancreatic intraepithelial neoplasia is the same as that of invasive pancreatic adenocarcinoma and cannot be used to differentiate between them. Pancreatic ductal adenocarcinomas resemble adenocarcinomas of the bile ducts and gallbladder in their light microscopy appearances and also in their immunophenotypical presentations. Pancreatic adenocarcinomas react with a variety of keratin antibodies, including CK8, CK17, CK18, CK19, CAM 5.2, and AE1/AE3.33,63 Pancreatic adenocarcinoma is generally both CK7 and CK20 positive. Pancreatic adenocarcinoma can also be faintly CDX2 positive.32,38 Almost all pancreaticobiliary adenocarcinomas are CEA positive and CA 125 positive. They may also have a minor component of neuroendocrine cells, which will react with somatostatin, synaptophysin, chromogranin, or other neuroendocrine markers.64 Mucin evaluation may be useful in evaluating prognosis.65,66 Therapy outcomes can also be predicted with positive vascular endothelial growth factor and negative SMAD4 (DPC4) immunostaining.60 Loss of expression of SMAD4 has also been shown in bile duct epithelium in cases of chronic gallstone disease.67 Well-differentiated metastatic pancreatic carcinoma to the liver may be difcult to distinguish from benign bile duct lesions in biopsy material. Unlike the benign lesions, however, they typically express p53, cytoplasmic mCEA, and other markers including CA 125.68 Neuroendocrine and Endocrine Cell Tumors, Low Grade and High Grade Low- and high-grade neuroendocrine tumors tend to show similar immunophenotypic expressions, but, in general, the intensity of staining is less with high-grade tumors. They can be grouped by the predominant secreted hormone (eg, somatostatin, gastrin) but usually also stain with synaptophysin and chromogranin, as well as with various keratins, including CK8, CK18, and CAM 5.2. CK7 and CK20 are generally negative.6973 CD56 and CD57 tend to stain more intensely in high-grade neuroendocrine tumors than in low grade, in a membranous pattern; CD56 will also be positive in a variety of other tumors.69,74 Serotonin can also be demonstrated.75 High-grade neuroendocrine tumors can also stain with calcitonin, and metastatic lesions can be misinterpreted as having arisen in the thyroid.73 Epithelial cytoplasmic expression of CD10, in contrast to membrane staining, is more commonly seen in malignant than in benign pancreatic endocrine tumors.76 Undifferentiated Solid Tumors of the Pancreas Undifferentiated Tumors Including Pleomorphic Carcinoma. These tumors can be complex, with adenocarcinoma and squamous differentiation along with spindle and undifferentiated areas. They can include rhabdoid elements. Patchy but strong immunostaining is seen with CK7 and CK19, with considerable variation in staining by most other keratins, including AE1/AE3, CAM 5.2, CK8, and CK18 (Table 7). Vimentin is usually positive.77 Osteoclast-like giant cells are often seen as a stromal component in undifferentiated carcinomas and can be immunoImmunohistochemistry of Liver and GI NeoplasmsGeller et al

CK7 CK20 CEA (monoclonal) CDX-2 TTF-1 ER / CA 125 Vimentin / / * CK indicates cytokeratin; , negative ( 10%); , positive ( 10%); CEA, carcinoembryonic antigen; TTF-1, thyroid transcription factor 1; ER, estrogen receptor; and / , equivocal.

Prognostic Markers In recent years, a number of immunohistochemical and molecular markers have been developed to predict outcome and also to help select therapies. Colon adenocarcinoma, the most frequent of malignant epithelial tumors affecting the gastrointestinal system, has been particularly evaluated in this way. Currently useful markers that can be studied with immunohistochemical techniques include evaluation of p27, p53, thymidylate synthase,56 and EGFR,57 although the usefulness of these markers, with the exception of EGFR, is still in question. The absence of p27 has been suggested to be a strong negative prognostic marker, particularly in stage II colon cancer, and may help select patients who will benet from adjuvant therapy. p53 nuclear expression is also associated with shortened survival and might be useful as an independent predictor in patients in whom colorectal carcinoma is metastatic to regional lymph nodes. Overexpression of thymidylate synthase has also been associated with poor prognosis and resistance to 5-ourouracil chemotherapy. EGFR expression is also an indicator of increased likelihood of metastases and decreased survival, but immunoreactivity is also associated with clinical response to cetuximab (Erbitux) therapy (Figure 5).57 Loss of expression of SMAD4 (DPC4), a gene found on chromosome 4, is also associated with poor prognosis in colorectal carcinoma.58 This immunostain is not diagnostically useful because it is also expressed in other carcinomas, including pancreas.59,60 Microsatellite instability can also be evaluated with immunohistochemical reactions that evaluate the proteins MLH-1, MSH-2, and MSH-6.61,62 Mutations in one or several DNA mismatch repair genes lead to the development of microsatellite instability (MSI) in hereditary nonpolyposis colorectal cancer and in 15% to 20% of patients with sporadic colorectal adenocarcinoma. Microsatellite instability is associated with improved survival in both sporadic and hereditary tumors. Mutations in mismatch repair genes result in absence of expression of one or more of the proteins; this is seen as negative immunostaining and is strongly associated with MSI. The presence of immunostaining is evidence of normal expression of the protein and shows that the mutation has not occurred. In this regard, immunohistochemistry is a sensitive ( 90%) and extremely specic (100%) method for screening for DNA mismatch repair defects. The predictive value for the absence of MSI with normal immunohistochemistry results is 97%, and the predictive value
496 Arch Pathol Lab MedVol 132, March 2008

Table 7.

Immunophenotype of Solid Tumors of the Pancreas*

Endocr Neo Acinar Cell Carcinoma Solid-Pseudopapillary Neoplasm

Immunostain

Keratin / (focal and faint) Vimentin Synaptophysin / (few cells) / Chromogranin / (few cells) Chymotrypsin Trypsin -Catenin / * Endocr Neo indicates endocrine neoplasm; , positive; / , equivocal ( 10% positive); and , negative or only rare cells.

prove to be prognostically valuable in predicting both tumor behavior and response to therapy.97 Functional proteomics, already used in the study of a variety of inammatory liver diseases to evaluate gene expression and protein production, will soon leave the research arena and will nd wide use in evaluating gene functions and molecular pathways integral to the progression of tumors.98,99 ADDITIONAL RESOURCES Three relatively recent texts may be useful as general references sources.100102 In addition, there is a Web-based reference (ImmunoQuery) that is frequently updated and provides literature references as well as data about immunostains and diagnoses.103
References
1. Taylor CR, Shi S-R, Barr NJ, Wu N. Techniques of immunohistochemistry: principles, pitfalls and standardization. In: Dabbs D, ed. Diagnostic Immunohistochemistry. 2nd ed. Philadelphia, Pa: Churchill Livingstone Elsevier; 2006:142. 2. Yaziji H, Barry T. Diagnostic immunohistochemistry: what can go wrong? Adv Anat Pathol. 2006;13:238246. 3. Lau SK, Prakash S, Geller SA, Alsabeh R. Comparative immunohistochemical prole of hepatocellular carcinoma, cholangiocarcinoma, and metastatic adenocarcinoma. Hum Pathol. 2002;33:11751181. 4. Supriatna Y, Kishimoto T, Uno T, Nagai Y, Ishikura H. Evidence for hepatocellular differentiation in alpha-fetoprotein-negative gastric adenocarcinoma with hepatoid morphology: a study with in situ hybridization for albumin mRNA. Pathology. 2005;37:211215. 5. Guzman G, Alagiozian-Angelova V, Layden-Almer JE, et al. p53, Ki-67, and serum alpha feto-protein as predictors of hepatocellular recurrence in liver transplant patients. Mod Pathol. 2005;18:14981503. 6. Jing Z, Nan KJ, Hu ML. Cell proliferation, apoptosis and the related regulators p27, p53 expression in hepatocellular carcinoma. World J Gastroenterol. 2005;11:19101916. 7. Zhang T, Sun HC, Xu Y, et al. Overexpression of platelet-derived growth factor receptor alpha in endothelial cells of hepatocellular carcinoma associated with high metastatic potential. Clin Cancer Res. 2005;11:85578563. 8. Van Eyken P, Sciot R, Brock P, et al. Abundant expression of cytokeratin 7 in brolamellar carcinoma of the liver. Histopathology. 1990;17:101107. 9. Berman MA, Burnham JA, Sheahan DG. Fibrolamellar carcinoma of the liver: an immunohistochemical study of nineteen cases and a review of the literature. Hum Pathol. 1988;19:784794. 10. Wang JH, Dhillon AP, Sankey EA, et al. Neuroendocrine differentiation in primary neoplasms of the liver. J Pathol. 1991;183:6167. 11. Buckley AF, Burgart LJ, Kakar S. Epidermal growth factor receptor expression and gene copy number in brolamellar hepatocellular carcinoma. Hum Pathol. 2006;37:410414. 12. Pan CC, Chen PC, Tsay SH, Ho DM. Differential immunoproles of hepatocellular carcinoma, renal cell carcinoma and adrenocortical carcinoma: a systemic immunohistochemical survey using tissue array technique. Appl Immunohistochem Mol Morphol. 2005;13:347352. 13. Stroescu C, Herlea V, Dragnea A, Popescu I. The diagnostic value of cytokeratins and carcinoembryonic immunostaining in differentiation hepatocellular carcinomas from intrahepatic cholangiocarcinomas. J Gastrointest Liver Dis. 2006;15:914. 14. Lei JY, Bourne PA, diSantAgnese PA, Huang J. Cytoplasmic staining of TTF-1 in the differential diagnosis of hepatocellular carcinoma vs. cholangiocarcinoma and metastatic carcinoma of the liver. Am J Clin Pathol. 2006;125:519 525. 15. Lodi C, Szabo E, Holczbauer A, et al. Claudin-4 differentiates biliary tract cancer from hepatocellular carcinomas. Mod Pathol. 2006;19:460469. 16. Loy TS, Quesenberry JT, Sharp SC. Distribution of CA 125 in adenocarcinomas: an immunohistochemical study of 481 cases. Am J Clin Pathol. 1992;98: 175179. 17. Paulsen FP, Varoga D, Paulsen AR, Coreld A, Tsokos M. Prognostic value of mucins in the classication of ampullary carcinomas. Hum Pathol. 2006;37: 160167. 18. Aishima S, Kuroda Y, Nishihara Y, et al. Gastric mucin phenotype denes tumor progression and prognosis of intrahepatic cholangiocarcinoma: gastric foveolar type is associated with aggressive tumor behavior. Histopathology. 2006; 49:3544. 19. Cazals-Hatem D, Rebouissou S, Bioulac-Sage P, et al. Clinical and molecular analysis of combined hepatocellular cholangiocarcinoma. J Hepatol. 2004; 41:292298. 20. Wakasa T, Wakasa K, Shutou T, et al. A histopathological study of combined hepatocellular and cholangiocarcinoma: cholangiocarcinoma component is originated from hepatocellular carcinoma. Hepatogastroenterology. 2007;54: 508513. 21. Hong SM, Cho H, Moskaluk CA. CDX2 and MUC2 protein expression in extrahepatic bile duct carcinoma. Am J Clin Pathol. 2005;124:361370. 22. Xu AM, Zhang SH, Zheng JM, Zheng WQ, Wu MC. Pathological and mo-

phenotypically similar to giant cell tumors of bone and may demonstrate K-ras mutations.78,79 They also stain with CD45, CD68, CD71, vimentin, and 1-antitrypsin.8082 Pancreatic osteoclast-like giant cell tumor can be present in association with mucus-secreting adenocarcinoma.83,84 Acinar Cell Carcinoma. Acinar cell carcinoma typically stains for enzymes including trypsin, chymotrypsin, lipase, and elastase (Figure 6). Pancreatic amylase is rarely detected in acinar cell carcinoma despite its presence in nonneoplastic acinar cells. Acinar cell carcinomas also commonly have focal endocrine differentiation. Endocrine differentiation in more than 25% of the neoplastic cells is classied as mixed acinar-endocrine carcinoma.8589 Most acinar cell carcinomas also stain with many keratins, including AE1/AE3, CAM 5.2, CK7, CK8, CK16, and CK19. Indeed, CEA negativity can be helpful to distinguish these tumors from the usual duct adenocarcinomas of the pancreas.89 Solid-Pseudopapillary Neoplasms. In general, these neoplasms of the pancreas do not react with most keratin antibodies and generally do not have neuroendocrine differentiation.90 Vimentin is generally positive as is 1-antitrypsin. Nuclear labeling with -catenin is observed in more than 90% of the tumors.91 Mucinous Cystic Neoplasms of the Pancreas and Liver. These tumors of the pancreas and liver can show stroma indistinguishable from that of the ovary, and immunostains for estrogen and progesterone receptors and CD10 may be positive.92 Intraductal Papillary Mucinous Neoplasms. Mucin expression may be useful in differentiating benign from malignant intraductal papillary-mucinous tumors of the pancreas and in predicting the potential for invasion.9395 MUC1 immunostaining has been shown to be a sensitive and specic indicator of invasiveness in intraductal papillary mucinous tumor.95 SMAD4 (DPC4) and p53 can also be useful but with lesser sensitivity. The claudin family of proteins can also help classify this group of neoplasms.96 FUTURE DIRECTIONS Immunohistochemical and molecular studies will be increasingly used in the study of gastrointestinal tumors, using techniques that are now only available in large medical centers. There will be an increasing number and more specic antibodies for immunoperoxidase and in situ hybridization studies. Fluorescence in situ hybridization, currently widely used in the study of leukemias, will also be more widely available and will be used for the study of many solid tumors. Cytogenetic studies have already been demonstrated to be of scientic interest and will likely
Arch Pathol Lab MedVol 132, March 2008

Immunohistochemistry of Liver and GI NeoplasmsGeller et al

497

lecular analysis of sporadic hepatic angiomyolipoma. Hum Pathol. 2006;37:735 741. 23. Yeh CN, Lee KF, Chen MF. Immunohistochemical study of hepatic angiomyolipoma. Hepatogastroenterology. 2005;52:11511153. 24. Cho NH, Shim HS, Choi YD, et al. Estrogen receptor is signicantly associated with the epithelioid variants of renal angiomyolipoma: a clinicopathologic and immunohistochemical study of 67 cases. Pathol Int. 2004;54:510515. 25. Makhlouf HR, Abdul-Al HM, Goodman ZD. Diagnosis of focal nodular hyperplasia of the liver by needle biopsy. Hum Pathol. 2005;36:12101216. 26. Nguyen BN, Flejou JF, Terris B, Belghiti J, Degott C. Focal nodular hyperplasia of the liver: a comprehensive pathologic study of 305 lesions and recognition of new histologic forms. Am J Surg Pathol. 1999;23:14411454. 27. Zucman-Rossi J, Jeannot E, Tran Van Nhieu J, et al. Genotype-phenotype correlation in hepatocellular adenoma: new classication and relationship with HCC. Hepatology. 2006;43:515524. 28. Bioulac-Sage P, Balabaud C, Bedossa P, et al. Pathological diagnosis of liver cell adenoma and focal nodular hyperplasia: Bordeaux update. J Hepatol. 2007;46:521527. 29. Chen ZM, Crone KG, Watson MA, Pfeifer JD, Wang HL. Identication of a unique gene expression signature that differentiates hepatocellular adenoma from well-differentiated hepatocellular carcinoma. Am J Surg Pathol. 2005;29: 16001608. 30. Libbrecht L, Severi T, Cassiman D, et al. Glycipan-3 expression distinguishes small hepatocellular carcinomas from cirrhosis, dysplastic nodules, and focal nodular hyperplasia-like nodules. Am J Surg Pathol. 2006;30:14051411. 31. Tot T. Adenocarcinomas metastatic to the liver: the value of cytokeratins 20 and 7 in the search for unknown primary tumors. Cancer. 1999;85:171177. 32. Tot T. Identifying colorectal metastases in liver biopsies: the novel CDX2 antibody is less specic than the cytokeratin 20 /7- phenotype. Med Sci Monit. 2004;10:BR139BR143. 33. Wang NP, Zee S, Zarbo RJ, et al. Coordinate expression of cytokeratins 7 and 20 denes unique subsets of carcinomas. Appl Immunohistochem. 1995;3: 99107. 34. Rullier A, Le Bail B, Fawaz R, et al. Cytokeratin 7 and 20 expression in cholangiocarcinomas varies along the biliary tract but still differs from that in colorectal carcinoma metastases. Am J Surg Pathol. 2000;24:870876. 35. Kende AI, Carr NJ, Sobin LH. Expression of cytokeratins 7 and 20 in carcinomas of the gastrointestinal tract. Histopathology. 2003;42:11371140. 36. Park SY, Kim HS, Hong EK, et al. Expression of cytokeratin 7 and 20 in carcinomas of the stomach and colorectum and their value in the differential diagnosis of metastatic carcinomas to the ovary. Hum Pathol. 2002;33:1078 1085. 37. Chu P, Wu E, Weiss LM. Cytokeratin 7 and cytokeratin 20 expression in epithelial neoplasms: a survey of 435 cases. Mod Pathol. 2000;13:962972. 38. Werling RW, Yaziji H, Bacchi CE, et al. CDX2, a highly sensitive and specic marker of adenocarcinomas of intestinal origin: an immunohistochemical survey of 476 primary and metastatic carcinomas. Am J Surg Pathol. 2003;27: 303310. 39. Moskaluk CA, Zhang H, Powell SM, et al. CDX2 protein expression in normal and malignant human tissues: immunohistochemical survey using tissue microarrays. Mod Pathol. 2003;16:913919. 40. Ha Kim G, Am Song G, Youn Park D, et al. CDX2 expression is increased in gastric cancers with less invasiveness and intestinal mucin phenotype. Scand J Gastroenterol. 2006;41:880886. 41. Roessler K, Monig SP, Schneider PM, et al. Co-expression of CDX2 and MUC2 in gastric carcinomas: correlation with clinico-pathologic parameters and prognosis. World J Gastroenterol. 2005;11:31823188. 42. Fan Z, Montgomery K, Rouse RV. Hep par 1 antibody stain for the differential diagnosis of hepatocellular carcinoma: 676 tumors tested using tissue microarrays and conventional tissue sections. Mod Pathol. 2003;16:137144. 43. Leteurtre E, Zerimech F, Piessen G, et al. Relationships between mucinous gastric carcinoma, MUC2 expression and survival. World J Gastroenterol. 2006; 7:33243331. 44. Burke AP, Thomas RM, Elsayed AM, et al. Carcinoids of the jejunum and ileum: an immunohistochemical and clinicopathologic study of 167 cases. Cancer. 1997;79:10861093. 45. Park JG, Choe GY, Helman LJ, et al. Chromogranin-A expression in gastric and colon cancer tissues. Int J Cancer. 1992;51:189194. 46. Boo YJ, Park SS, Kim JH, et al. Gastric neuroendocrine carcinoma: clinicopathologic review and immunohistochemical study of E-cadherin and Ki-67 as prognostic markers. J Surg Oncol. 2007;95:110117. 47. Blumenfeld W, Chandhoke DK, Sagerman P, et al. Neuroendocrine differentiation in gastric adenocarcinomas: an immunohistochemical study. Arch Pathol Lab Med. 1996;120:478481. 48. Wong NA, Young R, Malcomson RD, et al. Prognostic indicators for gastrointestinal stromal tumours: a clinicopathologic and immunohistochemical study of 108 resected cases of the stomach. Histopathology. 2003;43:118126. 49. Greenson JK. Gastrointestinal stromal tumors and other mesenchymal tumors of the gut. Mod Pathol. 2003;16:366375. 50. Goldblum JR. Gastrointestinal stromal tumors: a review of characteristic morphologic, immunohistochemical, and molecular genetic features. Am J Clin Pathol. 2002;117(suppl):S49S61. 51. de Silva CM, Reid R. Gastrointestinal stromal tumors (GIST): C-kit mutations, CD117 expression, differential diagnosis and targeted cancer therapy with imatinib. Pathol Oncol Res. 2003;9:1319.

52. Lee MJ, Lee HS, Kim WH, et al. Expression of mucins and cytokeratins in primary carcinomas of the digestive system. Mod Pathol. 2003;16:403410. 53. Albarracin CT, Jafri J, Montag AG, et al. Differential expression of MUC2 and MUC5AC mucin genes in primary ovarian and metastatic colon carcinoma. Hum Pathol. 2000;31:672677. 54. Chou YY, Jeng YM, Kao HL, et al. Differentiation of ovarian mucinous carcinoma and metastatic colorectal adenocarcinoma by immunostaining with beta-catenin. Histopathology. 2003;43:151156. 55. Nishizuka S, Chen ST, Gwadry FG, et al. Diagnostic markers than distinguish colon and ovarian adenocarcinomas: identication by genomic, proteomic, and tissue array proling. Cancer Res. 2003;63:52435250. 56. Compton C, Fenoglio-Presider C, Pettigrew N, Fielding L. American Joint Committee on Cancer prognostic factors consensus conference. Cancer. 2000; 88:17391757. 57. Cunningham D, Humblet Y, Siena S, et al. Cetuximab monotherapy and cetuximab plus irinotecan in irinotecan-refractory metastatic colorectal cancer. N Engl J Med. 2004;351:337345. 58. Isaksson-Mettavainio M, Palmqvist R, Forssell J, Stenling R, Oberg A. SMAD4/DPC4 expression and prognosis in human colorectal cancer. Anticancer Res. 2006;26:507510. 59. Sheehan GM, Kallakury BV, Sheehan CE, et al. Smad4 protein expression correlates with grade, stage, and DNA ploidy in prostatic adenocarcinomas. Hum Pathol. 2005;36:12041209. 60. Khorana AA, Hu YC, Ryan CK, et al. Vascular endothelial growth factor and DPC4 predict adjuvant therapy outcomes in resected pancreatic cancer. J Gastrointest Surg. 2005;9:903911. 61. Lindor N, Burgart L, Leontovich O, et al. Immunohistochemistry versus microsatellite instability testing in phenotyping colorectal tumors. J Clin Oncol. 2002;20:10431048. 62. Ward RL, Turner J, Williams R, et al. Routine testing for mismatch repair deciency in sporadic colorectal cancer is justied. J Pathol. 2006;208:590591. 63. Yamaguchi K, Enjoji M. Carcinoma of the pancreas: a clinicopathologic study of 96 cases with immunohistochemical observation for CEA and CA19-9. J Surg Oncol. 1991;47:148154. 64. Tamioloakis D, Simopoulos C, Venizolos J, et al. Distribution of somatostatin in pancreatic ductal adenocarcinoma remodels the normal pattern of the protein during foetal pancreatic development: an immunohistochemical analysis. Clin Exp Med. 2005;5:106111. 65. Kigure S. Immunohistochemical study of the association between the progression of pancreatic ductal lesions and the expression of MUC1, MUC2, MUC5AC, and E-cdherin. Rinsho Byori. 2006;54:447452. 66. Saitou M, Goto M, Horinouchi M, et al. MUC4 expression is a novel prognostic factor in patients with invasive ductal carcinoma of the pancreas. J Clin Pathol. 2005;58:845852. 67. Lee KT, Chang WT, Wang SN, et al. Expression of DPC4/Smad4 gene in stone-containing intrahepatic bile duct. J Surg Oncol. 2006;15:94:338343. 68. Hornick JL, Lauwers GY, Odze RD. Immunohistochemistry can help distinguish metastatic pancreatic adenocarcinomas from bile duct adenomas and hamartomas of the liver. Am J Surg Pathol. 2005;29:381389. 69. Mertz H, Vyberg M, Paulsen SM, et al. Immunohistochemical detection of neuroendocrine markers in tumors of the lungs and gastrointestinal tract. Appl Immmunohistochem. 1998;6:175180. 70. Bordi C, Yu JY, Baggi MT, et al. Gastric carcinoids and their precursor lesions: a histologic and immunohistochemical study of 23 cases. Cancer. 1991; 67:663672. 71. Lam KY, Lo CY. Pancreatic endocrine tumour: a 22-year clinico-pathological experience with morphological, immunohistochemical observation and a review of the literature. Eur J Surg Oncol. 1997;23:3642. 72. Rindi G, Paolotti D, LaRosa S, et al. The tumours of the endocrine pancreas. Eur J Histochem. 1998;42:6366. 73. Chetty R, Asa SL. Pancreatic endocrine tumors. Adv Anat Pathol. 2004;22: 202210. 74. Shipley WR, Hammer RD, Lennington WJ, et al. Parafn immunohistochemical detection of CD56, a useful marker for neural cell adhesion molecule (NCAM), in normal and neoplastic xed tissues. Appl Immunohistochem. 1997; 5:8793. 75. Soga J. Carcinoids of the pancreas: an analysis of 156 cases. Cancer. 2005; 104:11801187. 76. Deschamps L, Handra-Luca A, OToole D, et al. CD10 expression in pancreatic endocrine tumors: correlation with prognostic factors and survival. Hum Pathol. 2006;37:802808. 77. Paal E, Thompson LD, Frommelt RA, et al. A clinicopathological and immunohistochemical study of 35 anaplastic carcinomas of the pancreas with a review of the literature. Ann Diagn Pathol. 2001;5:129140. 78. Goldberg RD, Michelassi F, Montag AG. Osteoclast-like giant cell tumor of the pancreas: immunophenotypic similarity to giant cell tumor of bone. Hum Pathol. 1991;22:618622. 79. Westra WH, Sturm P, Drillenburg P, et al. K-ras oncogene mutations in osteoclast-like giant cell tumors of the pancreas and liver: genetic evidence to support origin from the duct epithelium. Am J Surg Pathol. 1998;22:12471254. 80. Imai Y, Morishita S, Ikeda Y, et al. Immunohistochemical and molecular analysis of giant cell carcinoma of the pancreast: a report of three cases. Pancreas. 1999;18:308315. 81. Gatteschi B, Saccomanno S, Bartoli FG, et al. Mixed pleomorphic-osteo-

498 Arch Pathol Lab MedVol 132, March 2008

Immunohistochemistry of Liver and GI NeoplasmsGeller et al

clastic-like tumor of the pancreas: light microscopical, immunohistochemical, and molecular biologic studies. Int J Pancreatol. 1995;18:169175. 82. Lucas Z, Dvorak K, Kroupova I, Valaskova I, Habanec B. Immunohistochemical and genetic analysis of osteoclastic giant cell tumor of the pancreas. Pancreas. 2006;32:325329. 83. Nai GA, Amico E, Giminez VR, Guilmar M. Osteoclast-like giant cell tumor of the pancreas associated with mucus-secreting adenocarcinoma: case report and discuss of the histogenesis. Pancreatology. 2005;5:279284. 84. Sedivy R, Kalipciyan M, Mazal PR, et al. Osteoclast-like giant cell tumor in mucinous cystadenocarcinoma of the pancreas: an immunohistochemical and molecular analysis. Cancer Detect Prev. 2005;29:814. 85. Klimstra DS, Longnecker D. Acinar cell carcinoma. In: Hamilton SR, Aaltonen LA, eds. Pathology and Genetics of Tumours of the Digestive System. Lyon, France: IARC Press; 2000. World Health Organization Classication of Tumours. 86. Fabre A, Sauvanet A, Flejou JF, et al. Intraductal acinar cell carcinoma of the pancreas. Virchows Arch. 2001;438:312315. 87. Ordonez NG. Pancreatic acinar cell carcinoma. Adv Anat Pathol. 2001;8: 144159. 88. Hartman GG, Ni H, Pickleman J. Acinar cell carcinoma of the pancreas. Arch Pathol Lab Med. 2001;125:11271128. 89. Klimstra DS, Heffess CS, Oertel JE, et al. Acinar cell carcinoma of the pancreas: a clinicopathologic study of 28 cases. Am J Surg Pathol. 1992;16:815 837. 90. Pettinato G, Di Vizio D, Manivel JC, et al. Solid-pseudopapillary tumor of the pancreas: a neoplasm with distinct and highly characteristic cytological features. Diagn Cytopathol. 2002;27:325334. 91. Abraham SC, Klimstra DS, Wilentz RE, et al. Solid-pseudopapillary tumors of the pancreas are genetically distinct from pancreatic ductal adenocarcinomas and almost always harbor B-catenin mutations. Am J Pathol. 2002;160:1361 1369. 92. Weihing RR, Shintaku IP, Geller SA, Petrovic LM. Hepatobiliary and pancreatic mucinous cystadenocarcinomas with mesenchymal stroma: analysis of

estrogen receptors/progesterone receptors and expression of tumor-associated antigens. Mod Pathol. 1997;10:372374. 93. Ito H, Endo T, Oka T, et al. Mucin expression prole is related to biological and clinical characteristics of intraductal papillary-mucinous tumors of the pancreas. Pancreas. 2005;30:96102. 94. Handra-Luca A, Flejou JF, Rufat P, et al. Human pancreatic mucinous cystadenoma is characterized by a distinct mucin, cytokeratin and CD10 expression compared with intraductal papillary-mucinous adenoma. Histopathology. 2006; 48:813821. 95. Ueda M, Miura Y, Kunihiro O, et al. MUC1 overexpression is the most reliable marker of invasive carcinoma in intraductal papillary-mucinous tumor (IPMT). Hepatogastroenterology. 2005;52:398403. 96. Tsukahara M, Nagai H, Kamiakito T, et al. Distinct expression patterns of claudin-1 and claudin-4 in intraductal papillary-mucinous tumors of the pancreas. Pathol Int. 2005;55:6369. 97. Diamond DL, Proll SC, Jacobs JM, et al. HepatoProteomics: applying proteomic technologies to the study of liver function and disease. Hepatology. 2006; 44:299308. 98. Chen R, Yi EC, Donohoe S, et al. Pancreatic cancer proteome: the proteins that underlie invasion, metastasis, and immunologic escape. Gastroenterology. 2005;129:11871197. 99. Finegold MJ. Hepatic tumors in childhood. In: Russo P, Ruchelli ED, Piccoli DA, eds. Pathology of Pediatric Gastrointestinal and Liver Disease. New York, NY: Springer; 2004:304313. 100. Dabbs D. Diagnostic Immunohistochemistry. 2nd ed. Philadelphia, Pa: Churchill Livingstone Elsevier; 2006. 101. Nadji M, Nassiri M, Morales AR. Efcient Tumor ImmunohistochemistryA Differential Diagnosis-Driven Approach. Chicago, Ill: ASCP Press; 2006. 102. Taylor CR, Cote RJ. ImmunomicroscopyA Diagnostic Tool for the Surgical Pathologist. 3rd ed. Philadelphia, Pa: WB Saunders; 2006. 103. ImmunoQuery, now STATdx PathIQ. Available at: https://mystatdxpathiq. com/PathIQ/Login.do. Accessed January 2008

Arch Pathol Lab MedVol 132, March 2008

Immunohistochemistry of Liver and GI NeoplasmsGeller et al

499