Glypican-3 as a useful diagnostic marker that distinguishes hepatocellular carcinoma from benign hepatocellular mass lesions.
Abstract: * Context.--Histopathologic distinction between hepatocellular carcinoma (HCC) and benign hepatocellular mass lesions, particularly hepatocellular adenoma, can sometimes be challenging. The currently available ancillary tools are suboptimal in terms of sensitivity and specificity.

Objective.--To further characterize the diagnostic value of glypican-3 (GPC3), a cell surface proteoglycan that has recently been shown to be overexpressed in HCC, in the distinction between HCC and benign hepatocellular mass lesions.

Design.--A total of 221 surgically resected liver specimens were subjected to immunohistochemical staining using a monoclonal antibody specific for GPC3. These included 111 HCCs, 48 hepatocellular adenomas, 30 focal nodular hyperplasias, and 32 large regenerative nodules in the background of cirrhosis.

Results.--Cytoplasmic, membranous, and canalicular staining for GPC3 was detected in 84 (75.7%) of the 111 HCCs, among which, 61 (72.6%) of the 84 cases exhibited diffuse immunoreactivity. In contrast, none of the 110 cases of hepatocellular adenoma, focal nodular hyperplasia, and large regenerative nodule showed detectable GPC3 staining. Focal GPC3 immunoreactivity was detected in cirrhotic nodules in 11 (16.4%) of 67 HCC cases with a cirrhotic background, but no background staining was observed in the remaining 44 HCCs without cirrhosis. GPC3 expression in HCCs did not correlate with the size, differentiation, or stage of the tumors; the presence or absence of cirrhotic background; or the underlying etiologies.

Conclusions.--GPC3 is a specific immunomarker for HCC that can be used to distinguish HCC from benign hepatocellular mass lesions, particularly hepatocellular adenoma. However, the diagnosis of HCC should not rely entirely on positive GPC3 immunostaining because focal immunoreactivity can be detected in a small subset of cirrhotic nodules. In addition, GPC3 expression in HCC can also be focal, and thus, the lack of GPC3 staining does not exclude the diagnosis of HCC.

(Arch Pathol Lab Med. 2008;132:1723-1728)
Article Type: Report
Subject: Hepatoma (Diagnosis)
Immunohistochemistry (Methods)
Proteoglycans (Health aspects)
Gene expression (Physiological aspects)
Genetic markers (Identification and classification)
Authors: Wang, Hanlin L.
Anatelli, Florencia
Zhai, Qihui "Jim"
Adley, Brian
Chuang, Shang-Tian
Yang, Ximing J.
Pub Date: 11/01/2008
Publication: Name: Archives of Pathology & Laboratory Medicine Publisher: College of American Pathologists Audience: Academic; Professional Format: Magazine/Journal Subject: Health Copyright: COPYRIGHT 2008 College of American Pathologists ISSN: 1543-2165
Issue: Date: Nov, 2008 Source Volume: 132 Source Issue: 11
Geographic: Geographic Scope: United States Geographic Code: 1USA United States
Accession Number: 230246857
Full Text: Glypican-3 (GPC3), a member of the heparan sulfate proteoglycan family, is an oncofetal protein that is expressed in the embryo and involved in morphogenesis and growth control during development. (1,2) Its expression is silenced in adult tissues, and loss-of-function mutations are responsible for Simpson-Golabi-Behmel syndrome, a rare X-linked prenatal and postnatal overgrowth with multiple congenital anomalies and increased risk of neoplasias in infancy. (3,4) In vitro studies have shown that GPC3 induces apoptosis in certain cell lines via the anchoring of the protein to the cell membrane, (5) indicating that GPC3 may function as an inhibitor of cell proliferation and a tumor suppressor in a cell line-specific manner.

Glypican-3 as a potential tumor marker for hepatocellular carcinoma (HCC) was first suggested in 1997 by Hsu et al, (6) who reported that GPC3 mRNA was preferentially expressed in HCCs. In their study of primary and/or recurrent HCCs from 154 patients, GPC3 (named MXR7 in their study) mRNA was detected in 143 (74.8%) of the 191 tumors, in contrast to only 3 cirrhotic and 2 (3.2%) of the noncirrhotic adult livers that showed low levels of GPC3 mRNA expression. In addition, GPC3 mRNA was not detected in 3 hepatocellular adenomas included in their study. These observations were subsequently confirmed by Zhu et al (7) in a similar study.

In 2003, several studies showed increased GPC3 expression at the protein level in HCC, when compared with healthy livers and benign hepatic lesions. (8-11) In the study by Capurro et al, (8) GPC3 protein expression was demonstrated by immunohistochemistry in 21 (72%) of 29 HCC cases but was not detected in healthy livers or various benign liver lesions, including 22 cirrhotic livers, 7 hepatocellular adenomas, and 4 focal nodular hyperplasias. In addition, GPC3 can be detected in the serum as a secreted protein in a subset of patients with HCC but is undetectable in healthy individuals or patients with hepatitis or cirrhosis. (8,10,12,13) Interestingly, elevated serum GPC3 levels in most HCC patients do not correlate with their serum [alpha]-fetoprotein values, and GPC3 appears to be more sensitive than [alpha]-fetoprotein as a serologic marker. Therefore, simultaneous use of both GPC3 and [alpha]-fetoprotein would improve overall sensitivity in HCC detection.

One of the challenges in histopathologic diagnosis of hepatocellular mass lesions is to distinguish HCC (particularly well differentiated) from hepatocellular adenoma, and the available immunomarkers in clinical practice, such as [alpha]-fetoprotein, polyclonal carcinoembryonic antigen, and CD34, have significant diagnostic limitations. (14,15) In this regard, GPC3 has shown promise as a useful diagnostic immunomarker as demonstrated in 3 studies in which a total of 28 hepatocellular adenomas were immunohistochemically examined. (8,16,17) In the current study, a large number of hepatocellular adenomas and other benign hepatocellular mass lesions were immunohistochemically analyzed for GPC3 expression to further characterize its diagnostic value in distinguishing them from HCC.

MATERIALS AND METHODS

A total of 221 liver resection specimens were included in this study. These included 111 HCCs, 48 hepatocellular adenomas, 30 focal nodular hyperplasias, and 32 large regenerative nodules arising in cirrhotic livers. These cases were retrieved from surgical pathology archives of Washington University Barnes-Jewish Hospital in St Louis, Mo; The Methodist Hospital in Houston, Tex; and Northwestern University Memorial Hospital in Chicago, Ill. The clinical history, pathology reports, and slides stained with hematoxylin-eosin were reviewed to confirm the diagnoses and to determine the underlying etiologies, when applicable. Hepatocellular carcinoma cases, in which the patients had undergone preoperative chemoembolization with tumor necrosis, were excluded from this study.

Immunohistochemical staining for GPC3 was performed following the protocol described previously. (18) Briefly, sections from formalin-fixed, paraffin-embedded tissue blocks were deparaffinized, rehydrated, and treated with 3% hydrogen peroxide for 15 minutes to inhibit endogenous peroxidase. Following microwave heat-induced epitope retrieval in 0.1M citrate buffer at pH 6.0 for 20 minutes, the slides were incubated with a mouse monoclonal antibody specific for GPC3 (clone 1G12) obtained from BioMosaics (Burlington, Vt), with a dilution of 1:200 for 1 hour at room temperature. After incubation with a rabbit, antimouse secondary antibody, a reaction was performed using the EnVision+ detection system that contains biotin-free horseradish peroxidase-labeled polymers obtained from Dako Corporation (Carpinteria, Calif). The staining was visualized using 3,3'-diaminobenzidine substrate-chromogen solution and counterstained with hematoxylin.

Immunohistochemically stained slides were independently evaluated by 3 observers (H.L.W., F.A., and X.J.Y.). Cases with significantly discrepant interpretation were resolved by rereview together by 2 of the observers (H.L.W. and F.A.). A case was considered negative if <5% of the cells of interest exhibited immunoreactivity. Positive stains were further stratified as focal (5% to 50% of the cells stained) and diffuse (>50% of the cells stained). Different staining patterns (cytoplasmic, membranous, or canalicular) were recorded.

Statistical analysis was performed using the 2-tailed Fisher exact test or the x2 test with the Yates continuity correction. A P value of <.05 was considered statistically significant.

RESULTS

Clinicopathologic Features of Hepatocellular Lesions

The ages of patients with HCC ranged from 26 to 82 years (mean, 60.1 years; median: 62 years). Of the 111 patients, 70 were men, and 41 were women, for a male to female ratio of 1.71:1. At the time of surgical resection, tumors ranged in size from 0.6 to 18.0 cm (mean, 5.2 cm; median, 4.0 cm). Tumors in 42 (37.8%) of the 111 cases

were stage I, 42 (37.8%) stage II, 19 (17.1%) stage IIIA, 2 (1.8%) stage IIIB, and 6 (5.4%) stage IV. Histopathologi cally, 51 (46%) of the HCC tumors were well differentiated, 45 (40.5%) were moderately differentiated, and 15 (13.5%) were poorly differentiated. Sixty-seven (60.4%) of the tumors were detected in a cirrhotic background, whereas 44 (39.6%) arose in noncirrhotic livers. The underlying etiologies associated with HCC development could not be determined in 33 cases (29.7%), among which, 78.8% (26/33) of the HCCs arose in noncirrhotic livers. In the remaining cases, hepatitis C was documented in 42 (37.8%) cases, hepatitis B in 12 cases (10.8%), hepatitis B and C coinfection in 2 cases (1.8%), hepatitis B and C coinfection plus excessive alcohol use in 6 cases (5.4%), excessive alcohol use in 9 cases (8.1%), possible nonalcoholic steatohepatitis in 5 cases (4.5%), familial hemochromatosis in 1 case (0.9%), and [[alpha].sub.1]-antitrypsin deficiency in 1 case (0.9%).

The ages of patients with hepatocellular adenoma ranged from 10 to 83 years (mean, 36.8 years; median, 33 years). Of the 48 patients, 43 (89.6%) were women and 5 (10.4%) were men, for a female to male ratio of 8.6:1. Eleven (25.6%) of the female patients had a documented history of using oral contraceptives. At the time of surgical resection, the tumors ranged in size from 0.6 to 25.0 cm (mean, 6.4 cm; median, 6.1 cm). Four (8.3%) of the 48 patients had more than 1 tumor.

The ages of patients with focal nodular hyperplasia ranged from 9 to 78 years (mean, 37.2 years; median, 38 years). Of the 30 patients, 28 (93.3%) were women and 2 (6.7%) were men, for a female to male ratio of 14:1. Two (7.1%) of the female patients had a documented history of using oral contraceptives. At the time of surgical resection, the lesions ranged in size from 0.3 to 14.0 cm (mean, 4.3 cm; median, 4.3 cm).

The ages of patients with large, regenerative nodules in a cirrhotic background ranged from 30 to 69 years (mean, 55.6 years; median, 58 years). Of the 32 patients, 21 (65.6%) were men and 11 (34.4%) were women, for a male to female ratio of 1.9:1. The underlying etiologies for cirrhosis included hepatitis C (14/32;43.8%), hepatitis B (2/ 32; 6.3%), hepatitis C and B coinfection (1/32; 3.1%), excessive alcohol use (3/32; 9.4%), a1-antitrypsin deficiency (2/32; 6.3%), primary sclerosing cholangitis (1/32; 3.1%), and cryptogenic origin (9/32; 28.1%).

Immunohistochemical Findings of GPC3 Expression in Hepatocellular Lesions

As shown in Table 1, positive immunostaining for GPC3 was observed in 84 (75.7%) of the 111 HCC cases, among which, 61 (72.6%) cases showed diffuse immunoreactivity. In marked contrast, none of the 110 benign hepatocellular mass lesions, including 48 hepatocellular adenomas, 30 focal nodular hyperplasias and 32 large regenerative nodules, showed GPC3 immunoreactivity.

In GPC3-expressing HCCs, 3 different immunostaining patterns were observed: predominantly membranous (Figure 1, A), predominantly canalicular (Figure 1, B), and predominantly cytoplasmic (Figure 1, C). These different staining patterns did not appear to correlate with the differentiation status or the growth pattern of the tumors. In fact, more than one third of the tumors exhibited a mixed staining pattern within the same tumors, although the tumor cells with different staining patterns showed similar or identical histologic features (Figure 2, A). Similarly, in those tumors that were focally positive for GPC3 expression, the tumor cells showing negative immunoreactivity did not appear to be dissimilar morphologically to those with positive staining within the same tumors (Figure 2, B).

Table 2 further demonstrates that GPC3 immunoreactivity detected in HCCs did not correlate with the differentiation status of the tumors. Although poorly differentiated tumors tended to express GPC3 more frequently than well-differentiated HCCs (93.3% vs 66.7%), the difference did not reach a statistical significance (P = .05). In addition, HCCs arising in cirrhotic livers expressed GPC3 at a similar frequency (76.1%) and with similar staining patterns (Figure 3, A) to those without a cirrhotic background (75%; Figure 3, B). Furthermore, no correlation was observed between GPC3 expression and the size or stage of the tumors or their underlying etiologies.

With respect to various morphologic variants, 5 (55.6%) of the 9 clear cell HCCs and 4 (80%) of 5 pelioid HCCs exhibited GPC3 immunoreactivity. There was only 1 case of sarcomatoid HCC included in our study, which showed focal GPC3 expression. There was also only 1 case of fibrolamellar HCC, which showed a diffuse staining pattern.

Among 67 HCCs with a cirrhotic background, focal GPC3 immunoreactivity was detected in a small proportion of the cirrhotic nodules in 11 (16.4%) cases. The staining typically involved periseptal hepatocytes, which were morphologically indistinguishable from the hepatocytes that did not stain for GPC3, either in the same nodules or in different nodules (Figure 4). The HCCs in these 11 cases were all positive for GPC3 expression, 7 (63.6%) diffuse and 4 (36.4%) focal, typically with a stronger staining intensity relative to that detected in cirrhotic nodules. No background staining was observed in nonneoplastic hepatocytes in any of the 44 HCC cases without cirrhosis.

COMMENT

HCC is the fifth most common cancer and the third most common cause of cancer-related deaths worldwide. (19) Accurate diagnosis is critically important to appropriate clinical management of the patients and assessment of the prognosis. Although histopathologic diagnosis of HCC may be straightforward in many cases, challenges are sometimes faced by pathologists in the distinction from benign hepatocellular mass lesions. This is particularly true of hepatocellular adenoma, which can sometimes be extremely difficult to distinguish from well differentiated HCC even on resection specimens. (14) It is unfortunate that the currently available immunomarkers, such as a-fetoprotein, polyclonal carcinoembryonic antigen, and CD34, have significant diagnostic limitations. (14,15)

[FIGURE 1 OMITTED]

[FIGURE 2 OMITTED]

[FIGURE 3 OMITTED]

In this study, we immunohistochemically examined a large number of hepatocellular mass lesions to further evaluate the diagnostic value of GPC3. Our findings extended the observations by others (8-11,16,17,20-22) and confirmed that GPC3 is a useful immunomarker to help separate HCC from hepatocellular adenoma and other benign hepatocellular mass lesions in a large cohort. Specifically, we detected GPC3 expression in 75.7% (84/111) of HCCs included in our study, a frequency comparable to those reported in previous studies. For example, in the study by Yamauchi et al, (16) positive GPC3 staining was observed in 47 (83.9%) of 56 HCCs. Interestingly, all positive cases in that study showed diffuse staining, which differs from the observations by us and others that positive cases frequently exhibit focal immunoreactivity. (8,20,22) This may have a considerable effect when using GPC3 immunostaining on needle core biopsies because a much lower detection rate should be expected because of sampling issues.23 It is interesting to note that heterogeneous GPC3 immunoreactivity and different staining patterns seen within the same tumor did not correlate with any morphologic characteristics.

Using a 1-mm tissue microarray, Wang et al (17) detected GPC3 expression in 38 (70.4%) of 54 HCCs. These authors also reported that GPC3 immunoreactivity was more frequently detected in HCCs with cirrhosis (19/21; 90%) than those without (18/28; 64%). However, this is not the case in our study where 51 (76.1%) of 67 HCCs with cirrhosis and 33 (75.0%) of 44 HCCs without cirrhosis were found to express GPC3. Wang et al (17) reported a P value of <.01 for the difference in their study, but recalculation of their data using the x2 test with the Yates continuity correction, as we did to our data, resulted in a P value of .08.

[FIGURE 4 OMITTED]

Our study also demonstrated that GPC3 expression in HCC did not correlate with the size or stage of the tumors or with the underlying etiologies. These findings are in accordance with previous observations. (16,20,21) It is somewhat unclear, however, whether GPC3 expression in HCC correlates with tumor differentiation. As observed by Yamauchi et al (16) and Libbrecht et al, (20) there appears to be a trend that well-differentiated HCCs express GPC3 less frequently than moderately and poorly differentiated HCCs, but the difference is not statistically significant. Nevertheless, in the study by Di Tommaso et al, (21) the authors reported that the number of GPC3-immunoreactive cells statistically increased with HCC dedifferentiation.

There have been only 3 prior studies, to our knowledge, that have examined GPC3 expression in a total of 28 hepatocellular adenomas. (8,16,17) In line with the observations in those studies, we found completely negative GPC3 staining in all 48 hepatocellular adenomas included in our study. It should be emphasized, however, that well-differentiated HCC, which poses the most diagnostic difficulties in the distinction from hepatocellular adenoma, frequently fails to express GPC3. In our series, one third of well-differentiated HCCs were completely negative for GPC3 expression, and another third showed only focal immunoreactivity. Therefore, GPC3 immunostaining is helpful only when it is positive, and negative staining should not be viewed as evidence to exclude the possibility of HCC. This is particularly true of needle core biopsies. (23) Interestingly, Wang et al (17) detected GPC3 expression in 3 (60%) of 5 atypical hepatocellular adenomas, suggesting that GPC3 could be helpful in identifying cases with malignant progression. However, malignant transformation appears to be a rare event because none of our adenoma cases showed any detectable GPC3 immunostaining.

Another interesting observation in our study is the detection of focal GPC3 expression in cirrhotic nodules in a small subset of HCC cases, but not in cirrhotic livers or even large regenerative nodules in cases without HCC. This raises the possibility that GPC3 may function as an early biomarker in hepatocarcinogenesis. In this regard, several studies have attempted to examine GPC3 expression in dysplastic nodules, a poorly defined premalignant nodular lesion arising in cirrhotic livers. (24) Overall, to our knowledge, a total of 160 dysplastic nodules (69 low grade, 77 high grade, 14 unspecified) have been examined by immunohistochemistry, and only 17 (10.6%) of the nodules showed positive staining, usually focal. (8,16,17,20-22) Although high-grade dysplastic nodules tended to express GPC3 more frequently (13/77; 16.9%) than low-grade dysplastic nodules (4/69; 5.8%), no statistically significant difference was detected (P = .07). It should be pointed out that the results reported by different investigators varied significantly. For example, Yamauchi et al (16) reported GPC3 expression in 2 (25%) of 8 low-grade and 6 (75%) of 8 high-grade dysplastic nodules. On the other hand, Libbrecht et al (20) showed no GPC3 immunostaining in 16 low-grade dysplastic nodules and focal staining in only 2 (6%) of 33 high-grade dysplastic nodules. One explanation for the variable results in different studies may be the difficulty and wide interobserver variability in the diagnosis of dysplastic nodules. (25) In our study, we intentionally excluded dysplastic nodules because we believed that the results of immunochemical studies based on inaccurate or disputable diagnoses would not be reliable or clinically useful. (25,26) Nevertheless, it is interesting to note that the GPC3-positive cirrhotic nodules in HCC cases in our series did not appear to exhibit dysplastic features and were morphologically indistinguishable from GPC3-negative nodules. Although the underlying mechanism(s) for selective GPC3 expression in cirrhotic nodules remains elusive, it is our speculation that the GPC3-expressing hepatocytes in cirrhotic nodules may have already initiated the neoplastic process at the molecular level. It is possible, therefore, that reactivation of GPC3 expression may represent an early molecular event in hepatocarcinogenesis, which precedes actual morphologic transformation.

In summary, our data demonstrate that GPC3 is a reliable diagnostic immunomarker, when positive, to distinguish HCC from benign hepatocellular mass lesions, including hepatocellular adenoma. Our findings emphasize that GPC3 immunoreactivity can be focal and heterogeneous, and thus negative immunostaining should not exclude the diagnosis of HCC. Our study also shows focal GPC3 expression in a small subset of cirrhotic nodules associated with HCCs, which should be taken into consideration when interpreting challenging needle biopsies.

Accepted for publication April 14, 2008.

References

(1.) Jakubovic BD, Jothy S. Glypican-3: from the mutations of Simpson-Golabi-Behmel genetic syndrome to a tumor marker for hepatocellular carcinoma. Exp Mol Pathol. 2007;82:184-1 89.

(2.) Li M, Choo B, Wong ZM, Filmus J, Buick RN. Expression of OCI-5/glypican 3 during intestinal morphogenesis: regulation by cell shape in intestinal epithelial cells. Exp Cell Res. 1997;235:3-1 2.

(3.) Pilia G, Hughes-Benzie RM, MacKenzie A, et al. Mutations in GPC3, a glypican gene, cause the Simpson-Golabi-Behmel overgrowth syndrome. Nat Genet. 1996;12:241-247.

(4.) Neri G, Gurrieri F, Zanni G, Lin A. Clinical and molecular aspects of the Simpson-Golabi-Behmel syndrome. Am J Med Genet. 1998;79:279-283.

(5.) Gonzalez AD, Kaya M, Shi W, et al. OCI-5/GPC3, a glypican encoded by a gene that is mutated in the Simpson-Golabi-Behmel overgrowth syndrome, induces apoptosis in a cell line-specific manner. J Cell Biol. 1998;141:1407-1414.

(6.) Hsu HC, Cheng W, Lai PL. Cloning and expression of a developmentally regulated transcript MXR7 in hepatocellular carcinoma: biological significance and temporospatial distribution. Cancer Res. 1997;57:5179-5184.

(7.) Zhu ZW, Friess H, Wang L, et al. Enhanced glypican-3 expression differentiates the majority of hepatocellular carcinomas from benign hepatic disorders. Gut. 2001;48:558-564.

(8.) Capurro M, Wanless IR, Sherman M, et al. Glypican-3: a novel serum and histochemical marker for hepatocellular carcinoma. Gastroenterology. 2003;125: 89-97.

(9.) Midorikawa Y, Ishikawa S, Iwanari H, et al. Glypican-3, overexpressed in hepatocellular carcinoma, modulates FGF2 and BMP-7 signaling. Int J Cancer. 2003;103:455-465.

(10.) Nakatsura T, Yoshitake Y, Senju S, et al. Glypican-3, overexpressed specifically in human hepatocellular carcinoma, is a novel tumor marker. Biochem Biophys Res Commun. 2003;306:16-25.

(11.) Sung YK, Hwang SY, Park MK, etal. Glypican-3 is overexpressed in human hepatocellular carcinoma. Cancer Sci. 2003;94:259-262.

(12.) Filmus J, Capurro M. Glypican-3 and alpha-fetoprotein as diagnostic tests for hepatocellular carcinoma. Mol Diagn. 2004;8:207-212.

(13.) Hippo Y, Watanabe K, Watanabe A, et al. Identification of soluble NH2-terminal fragment of glypican-3 as a serological marker for early-stage hepato cellular carcinoma. Cancer Res. 2004;64:2418-2423.

(14.) Chen ZM, Crone KG, Watson MA, Pfeifer JD, Wang HL. Identification of a unique gene expression signature that differentiates hepatocellular adenoma from well-differentiated hepatocellular carcinoma. Am J Surg Pathol. 2005;29: 1600-1608.

(15.) Wee A. Diagnostic utility of immunohistochemistry in hepatocellular carcinoma, its variants and their mimics. Appl Immunohistochem Mol Morphol. 2006;14:266-272.

(16.) Yamauchi N, Watanabe A, Hishinuma M, et al. The glypican 3 oncofetal protein is a promising diagnostic marker for hepatocellular carcinoma. Mod Pathol. 2005;18:1591-1598.

(17.) Wang XY, Degos F, Dubois S, et al. Glypican-3 expression in hepatocellular tumors: diagnostic value for preneoplastic lesions and hepatocellular carcinomas. Hum Pathol. 2006;37:1435-1441.

(18.) Zynger DL, Dimov ND, Luan C, Teh BT, Yang XJ. Glypican 3: a novel marker in testicular germ cell tumors. Am J Surg Pathol. 2006;30:1570-1575.

(19.) El-Serag HB. Hepatocellular carcinoma: an epidemiologic view. J Clin Gastroenterol. 2002;35:S72-S78.

(20.) Libbrecht L, Severi T, Cassiman D, etal. Glypican-3 expression distinguishes small hepatocellular carcinomas from cirrhosis, dysplastic nodules, and focal nodular hyperplasia-like nodules. Am J Surg Pathol. 2006;30:1405-1411.

(21.) Di Tommaso L, Franchi G, Park YN, et al. Diagnostic value of HSP70, glypican 3, and glutamine synthetase in hepatocellular nodules in cirrhosis. Hep atology. 2007;45:725-734.

(22.) Llovet JM, Chen Y, Wurmbach E, et al. A molecular signature to discriminate dysplastic nodules from early hepatocellular carcinoma in HCV cirrhosis. Gastroenterology. 2006;131:1758-1767.

(23.) Anatelli F, Chuang ST, Yang XJ, Wang HL. Value of glypican 3 immunostaining in the diagnosis of hepatocellular carcinoma on needle biopsy. Am J Clin Pathol. 2008;130:219-223.

(24.) International Working Party. Terminology of nodular hepatocellular lesions. Hepatology 1995;22:983-993.

(25.) Lennerz JK, Brunt EM, Dehner LP, et al. Observer variability in the histopathologic diagnosis of hepatocellular nodules [abstract]. Mod Pathol. 2008; 21(suppl 1):308.

(26.) Lennerz JK, Brunt EM, Dehner LP, et al. Immunoprofiling of hepatocellular nodules [abstract]. Mod Pathol. 2008;21(suppl 1):308-309.

Hanlin L. Wang, MD, PhD; Florencia Anatelli, MD; Qihui "Jim" Zhai, MD; Brian Adley, MD; Shang-Tian Chuang, DO; Ximing J. Yang, MD, PhD

From the Department of Pathology and Laboratory Medicine, Cedars-Sinai Medical Center, Los Angeles, Calif (Dr Wang); the Department of Pathology and Immunology, Washington University School of Medicine, St Louis, Mo (Dr Anatelli); the Department of Pathology, Weill Medical College of Cornell University, The Methodist Hospital, Houston, Tex (Dr Zhai); and the Department of Pathology, Northwestern University Feinberg School of Medicine, Chicago, Ill (Drs Adley, Chuang, and Yang).

The authors have no relevant financial interest in the products or companies described in this article.

Presented in part at the 96th Annual Meeting of the United States and Canadian International Academy of Pathology, San Diego, Calif, March 24-30, 2007.

Reprints: Hanlin L. Wang, MD, PhD, Department of Pathology and Laboratory Medicine, Cedars-Sinai Medical Center, 8700 Beverly Blvd, Los Angeles, CA 90048 (e-mail: hanlin.wang@cshs.org).
Table 1. Summary of Immunohistochemical Findings
of Glypican-3 Expression in Hepatocellular
Mass Lesions *

                   HCC,        HCA,       FNH,       LRN,
Immunoreactivity   No. (%)     No. (%)    No. (%)    No. (%)
(%)                (n = 111)   (n = 48)   (n = 30)   (n = 32)

Negative (<5)      27 (24.3)   48 (100)   30 (100)   32 (100)
Focal (5-50)       23 (20.7)   0          0          0
Diffuse (>50)      61 (55.0)   0          0          0

* HCC indicates hepatocellular carcinoma; HCA, hepatocellular
adenoma; FNH, focal nodular hyperplasia; and LRN, large regenerative
nodule.

Table 2. Correlation of Glypican-3 Expression in
Hepatocellular Carcinoma With Tumor Differentiation

                  No. (%) of cases
Tumor
Differentiation   Negative    Focal       Diffuse     Total

Well              17 (33.3)   15 (29.4)   19 (37.3)   51
Moderate           9 (20.0)    6 (13.3)   30 (66.7)   45
Poor               1 (6.7)     2 (13.3)   12 (80.0)   15
Total             27 (24.3)   23 (20.7)   61 (55.0)   111
Gale Copyright: Copyright 2008 Gale, Cengage Learning. All rights reserved.