Immunohistochemical Analysis of Nuclear Versus Cytoplasmic Staining of WT1 in Malignant Mesotheliomas and Primary Pulmonary Adenocarcinomas

Transcription

1 Immunohistochemical Analysis of Nuclear Versus Cytoplasmic Staining of WT1 in Malignant s and Primary Pulmonary s Matthew R. Foster, MD; Joyce E. Johnson, MD; Sandy J. Olson, MS; D. Craig Allred, MD Context. Previous studies have indicated certain immunohistochemical markers, including WT1, may be helpful in distinguishing adenocarcinomas from mesotheliomas, but to date there are no reliable, widely accepted, commercially available antibodies positive in mesotheliomas and negative in adenocarcinomas. We compared the nuclear and cytoplasmic staining patterns of WT1 in these 2 malignancies using a commercially available antibody and examined the expression of 2 other previously reported positive markers, calretinin and thrombomodulin. Methods. Sixty-seven mesotheliomas and 51 adenocarcinomas, all paraffin embedded, were retrieved from recent case files. The diagnosis of mesothelioma was based on typical clinical and morphologic features, as well as immunohistochemistry; electron microscopy had been performed on 16 cases. The diagnosis of adenocarcinoma was based on typical light microscopic findings and a positive stain for mucin. Commercially available antibodies to WT1, thrombomodulin, and calretinin were applied. Because of the conflict surrounding calretinin, 2 anticalretinin antibodies (from Chemicon Inc and Zymed Laboratories) were utilized. Results. Fifty of 67 mesotheliomas showed strong nuclear staining with WT1. No adenocarcinomas (0/51) showed nuclear staining. Twenty-three of 67 mesotheliomas were positive for thrombomodulin, and 35 of 67 mesotheliomas were positive for calretinin with the Chemicon antibody. Nine of 15 mesotheliomas were positive for calretinin with the Zymed antibody. Conclusions. Thrombomodulin and calretinin did not prove useful in discriminating between mesotheliomas and adenocarcinomas. The degree of positivity with calretinin may be dependent on the specific antibody utilized. Nuclear staining for WT1 is highly specific for mesothelioma and, in the appropriate clinical setting, can be a helpful adjunct in the distinction between adenocarcinomas and mesotheliomas. (Arch Pathol Lab Med. 2001;125: ) The distinction between some mesotheliomas and adenocarcinomas remains a diagnostic challenge, at times requiring a battery of immunohistochemical stains or electron microscopy for confirmatory diagnosis. Previous studies indicate certain immunohistochemical markers may be helpful in distinguishing adenocarcinomas from mesotheliomas, but no study has identified a marker that clearly and consistently makes this distinction in all cases. Many traditional markers, including carcinoembryonic antigen, Leu-M1, Moc-31, E-cadherin, thyroid transcription factor-1, and Ber-Ep4 are negative in mesotheliomas and positive in adenocarcinomas. Recent work has focused on other markers, including thrombomodulin, calretinin, CK5/6, n-cadherin, HBME-1, CD44S, and WT1, which are expressed in mesotheliomas However, to date there are no reliable, widely used, commercially available antibod- Accepted for publication May 24, From the Department of Pathology, Vanderbilt University Medical Center, Nashville, Tenn (Drs Foster, Johnson, and Olson); and the Breast Center, Baylor College of Medicine, Houston, Tex (Dr Allred). Presented in part at the Annual Meeting of the United States and Canadian Academy of Pathology, San Francisco, Calif, March 22, Reprints: Matthew R. Foster, MD, Department of Pathology, Vanderbilt University Medical Center, C3319 MCN, Nashville, TN ies that are positive in mesotheliomas and negative in adenocarcinomas. The Wilms tumor gene encodes a protein (WT1) that is expressed in the developing fetal kidneys and urogenital tract, the developing spleen, and in fetal coelomic lining cells, including mesothelium It is also expressed in mature benign mesothelial cells, and WT1 may play a role in the malignant transformation of certain cells These observations emphasize the potential of WT1 as a positive marker in mesotheliomas. Studies using noncommercial anti-wt1 antibodies have demonstrated discriminatory expression between mesotheliomas and adenocarcinomas. Further studies reported strong differential nuclear staining in mesotheliomas and cytoplasmic or no staining in adenocarcinomas. 25,26 However, to date there have been no studies comparing the nuclear and cytoplasmic staining patterns of these 2 malignancies using a commercially available antibody to WT1 on a relatively large number of cases. Using such an antibody, we report the nuclear and cytoplasmic expression of WT1 in malignant mesotheliomas and in primary pulmonary adenocarcinomas, as well as the expression of 2 other positive markers, calretinin and thrombomodulin Arch Pathol Lab Med Vol 125, October 2001 Immunohistochemical Analysis of WT1 Staining Foster et al

2 METHODS Sixty-seven mesotheliomas and 51 primary pulmonary adenocarcinomas were retrieved from the recent files of the Department of Surgical Pathology at Vanderbilt University Medical Center (17 mesotheliomas, 36 adenocarcinomas) and Saint Thomas Hospital (9 mesotheliomas, 15 adenocarcinomas) in Nashville, Tenn, using a SNOMED search for pleura (T29000), lung (T28000), adenocarcinoma (M81403), and mesothelioma (M90503). Additionally, 1 mesothelioma was retrieved from Baptist Hospital in Nashville and 40 mesotheliomas were obtained from The University of Texas Health Sciences Center, San Antonio, Tex. All cases were routinely fixed in 10% neutral buffered formalin for 12 to 24 hours at room temperature and embedded in paraffin. Routine 5- m sections were obtained from these blocks. Of the 67 mesotheliomas, 58 were considered epithelial, 6 sarcomatoid, and 3 biphasic. Additionally, 7 of 67 mesotheliomas were based in the peritoneum. The diagnosis of mesothelioma was based on the typical clinical and morphologic features, as well as immunohistochemistry; electron microscopy had been performed on 16 cases (15 cases from Vanderbilt University Medical Center and 1 case from Baptist Hospital). All adenocarcinomas were moderately to poorly differentiated and included bronchoalveolar and solid type with mucin production, as classified according to current World Health Organization (1999) criteria. The diagnosis of adenocarcinoma was based on the typical light microscopic findings, a positive stain for mucin, or both. Immunohistochemistry was carried out on an automated stainer (Ventana-ES, Ventana Medical Systems Inc, Tucson, Ariz) using standard techniques (biotin-peroxidase conjugated avidin method) and the following commercially available antibodies: a monoclonal antibody to thrombomodulin (Dako Corporation, Carpinteria, Calif, clone 1009, no pretreatment, 1:50 dilution); a polyclonal antibody raised in rabbits against guinea pig calretinin (AB149, Chemicon Inc, Temecula, Calif, microwave pretreatment in citrate buffer [ph 6.0], 1:100 dilution); and a polyclonal antibody against WT1 (Santa Cruz Biotechnology Inc, Santa Cruz, Calif, microwave pretreatment in citrate buffer [ph 6.0], 1: 100 dilution). Fetal kidney (WT1), tonsil (thrombomodulin), and adult testis (calretinin) were used as positive controls. Additionally, a second anticalretinin, a polyclonal antibody raised in rabbits against human calretinin (Zymed Laboratories, San Francisco, Calif, steamer pretreatment in citrate buffer [ph 6.0], prediluted by manufacturer) was applied to a selected group of 15 mesotheliomas and 15 adenocarcinomas. Both groups included a mix of positive and negative cases with the Chemicon anticalretinin antibody. Immunoreactivity was scored independently by 2 authors (M.F. and J.J.), who were blinded to the final diagnosis, according to percentage of immunoreactivity as follows: 25% of the malignant cells staining, 1 ; 25% to 50% staining, 2 ; 50% to 75% staining, 3 ; and 75% staining, 4. Additionally for WT1, the subcellular localization (cytoplasmic or nuclear) of any positive staining was noted. Table 1. Nuclear WT1 Cytoplasmic WT1 Thrombomodulin Calretinin Chemicon Zymed Results of Immunohistochemistry* 50/67 (75) 8/67 (12) 23/67 (34) 35/67 (52) 9/15 (60) 0/51 (0) 44/51 (86) 4/51 (7.8) 35/51 (68) 3/15 (20) * Values are expressed as the total number of cases marking with a given antibody (total positive/total number of cases) and the percent positive is shown in parentheses. P value for nuclear WT1 staining.001. Table 2. Differential Nuclear Versus Cytoplasmic Staining of s and s by WT1, and the Extent of Positivity Score Nuclear Cytoplasmic Extent* Arch Pathol Lab Med Vol 125, October 2001 Immunohistochemical Analysis of WT1 Staining Foster et al /50 8/ /50 0/8 Extent* 1 2 0/44 16/ /44 28/44 *1 indicates weakest ( 25%) and 4, strongest ( 75%). Based on 58 positive mesotheliomas and 44 positive adenocarcinomas. RESULTS Fifty-eight (86%) of 67 mesotheliomas showed positivity (cytoplasmic or nuclear) for WT1. As reflected in Table 1, strong nuclear staining with WT1 was observed in 50 (75%) of 67 of the mesotheliomas. Of the 50 cases showing positive nuclear staining, 35 (70%) showed strong staining, with greater than 50% and often greater than 75% of the neoplastic cells staining (Figure 1). Eight (12%) of 67 cases showed predominantly cytoplasmic staining, and in all 8 cases fewer than 50% of the cells demonstrated this pattern of staining (Table 2). Conversely, no adenocarcinomas (0/51) showed nuclear staining for WT1. Forty-four (86%) of 51 adenocarcinomas demonstrated variable cytoplasmic staining, and in 28 (63%) of 44 of these cases, greater than 50% of the neoplastic cells showed intermediate to strong cytoplasmic staining (Figure 2). Seven (14%) of 51 had no discernible positive stain. Twenty-three (34%) of 67 mesotheliomas showed positive staining for thrombomodulin, and only 8 (12%) of 67 demonstrated greater than 50% positivity. Most adenocarcinomas (47/51 [92%]) were negative for thrombomodulin. The remaining 4 cases had less than 50%, weak to intermediate staining patterns (Table 3). Using the Chemicon anticalretinin antibody, 35 (52%) of 67 mesotheliomas showed some positive staining for calretinin, and 30 (86%) of 35 demonstrated greater than 50% staining. Thirty-five of 51 adenocarcinomas showed at least some staining, and in 23 (45%) of 51, greater than 50% of the neoplastic cells marked with calretinin (Table 3). Using the Zymed anticalretinin antibody, 9 (60%) of 15 mesotheliomas showed some degree of positivity, and only 3 (20%) of 15 adenocarcinomas showed weak but definitive staining (Table 4). When stratified according to morphologic subtype, no significant difference was seen in nuclear WT1 expression, although the number of nonepithelial cases was small. As reflected in Table 5, 42 of 50 epithelial, 5 of 6 sarcomatoid, and 3 of 3 biphasic mesotheliomas showed nuclear expression of WT1. No sarcomatoid mesothelioma showed strong (4 ) staining with WT1; 4 showed weak (1 2 ) nuclear staining, and only 1 was graded as having 3 staining (Figure 3). Finally, although the number of peritoneal cases was small, differences in anatomic location (pleural vs peritoneal) showed no difference in antigen expression. Seven of 67 mesotheliomas were based in the peritoneum; 6 of 7 showed strong nuclear staining for WT1, 3 of 7 showed strong calretinin staining (Chemicon antibody), and only 1 of 7 were strongly positive for thrombomodulin. COMMENT In this series of cases, nuclear WT1 staining was 75% sensitive and 100% specific for mesotheliomas and dem-

3 1318 Arch Pathol Lab Med Vol 125, October 2001 Immunohistochemical Analysis of WT1 Staining Foster et al

4 3 4 * 1 2 * 0 (negative) Table 3. Differential Staining of Thrombomodulin and Calretinin and the Extent of Positivity Score 8/67 15/67 44/67 Thrombomodulin *1 indicates weakest ( 25%) and 4, strongest ( 75%). The results shown are using the Chemicon anticalretinin. 0/51 4/51 47/51 30/67 5/67 32/67 Calretinin 23/51 12/51 16/51 Table 4. Differential Results With 2 Different Commercially Available Anticalretinin Antibodies* s s Chemicon 35/67 (52)* 35/51 (68)* Zymed 9/15 (60)* 3/15 (20)* * Values presented as absolute numbers of positive cases per number tested (percent of positive). Epithelial Sarcomatoid Biphasic Table 5. Results of Staining Based on Morphologic Subtype Nuclear WT1 Calretinin* Thrombomodulin 42/50 5/6 3/3 31/58 2/6 2/3 20/58 2/6 1/3 * The results shown are using the Chemicon anticalretinin. onstrated a positive predictive value of 100% and a negative predictive value of 75%. Chi-square analysis of nuclear-positive versus nuclear-negative staining showed this distribution was significant with a P value less than.001 ( ). Additionally, cytoplasmic staining of the adenocarcinomas by WT1 was common (86%). Therefore, nuclear location of staining is critical for specificity. It remains unclear what precisely is showing antigenicity in those mesotheliomas and adenocarcinomas that demonstrated predominantly cytoplasmic staining with WT1. A cross-reactive, cytoplasmic antigen seems a plausible explanation for this observation. Indeed, investigators in prior studies also observed this phenomenon and proposed that this pattern may reflect reactivity to an epitope not related to WT1. 22,27 Others reported fixation may affect antigen detection. 28 The Wilms tumor gene, located on the short arm of chromosome 11 (11p13), encodes a protein with several functions, including as a potential tumor suppressor gene. 26 The detectable presence of WT1 in benign mesothelial cells suggests the expression of the Wilms tumor gene in mesothelial cells is appropriate and that its presence may be a function of cellular differentiation rather than a function of overexpression related to malignancy. However, its potential function as a tumor suppressor gene suggests a possible role in the malignant transformation of mesothelial cells. Our study confirms the expression of WT1 in mesotheliomas and more importantly highlights the specificity of nuclear staining in malignant mesotheliomas. Furthermore, the fact that no adenocarcinomas showed nuclear staining supports the hypothesis that WT1 may have some role in the malignant transformation of mesothelial cells. Although the main focus of this study was WT1 expression, the widely varied reports surrounding calretinin and thrombomodulin positivity remain an important component. Varying reports assert encouraging results with thrombomodulin and calretinin, but wisely caution that these markers may be a helpful addition to current practices and do not yet stand alone. 3,29 Others demonstrate strong results with calretinin and suggest this marker may be most useful in distinguishing epithelial-type mesotheliomas from adenocarcinomas. 5,8 However, prior work shows a distinct difference in the staining pattern of anticalretinin antibodies. Using a polyclonal anticalretinin manufactured by Zymed, Ordoñez reported strong differential expression in 38 of 38 mesotheliomas and using the anticalretinin antibody manufactured by Chemicon, reported differential staining in 28 of 38 cases. Thus, as previously reported, varying results may in fact be dependent on the specific anticalretinin antibody used in a given study. 2,30 Indeed, 5 studies using the anticalretinin from either Swant (Bellinzona, Switzerland) or from Zymed (San Francisco) have shown encouraging results, with reactivity (in some cases) seen in 100% of mesotheliomas and in less than 10% of adenocarcinomas. 5,8,31 33 Our initial results with the Chemicon antibody seem in stark contrast to these studies. As a result of the reported differences in commercial anticalretinins, we repeated our staining on a limited number of mesotheliomas and adenocarcinomas using a different commercially available anticalretinin antibody, and a method nearly identical to that used in previous studies. The results of our trial do not support the utility of calretinin as a differential marker and parallel those seen in a study by Riera et al, 34 which reported positivity for the Chemicon anticalretinin in 24 Figure 1. Low-magnification (A) and high-magnification (B) views of an epithelial malignant mesothelioma demonstrating 4 nuclear immunoperoxidase staining for WT1. Confirmatory electron microscopy had been performed previously on this case. C and D represent 2 cases of epithelial malignant mesotheliomas demonstrating 1 2 staining with WT1 (original magnification 100 [A], 400 [B], 100 [C], and 100 [D]). Figure 2. Low-magnification (A) and high-magnification (B) views of a moderately differentiated primary pulmonary adenocarcinoma demonstrating 3 4 cytoplasmic staining for WT1. No adenocarcinomas showed nuclear staining for WT1 (original magnification 100 [A] and 400 [B]). Figure 3. Low-magnification view of 2 different sarcomatoid mesotheliomas, each showing 2 nuclear staining with WT1 (original magnification 100 [A] and B]). Arch Pathol Lab Med Vol 125, October 2001 Immunohistochemical Analysis of WT1 Staining Foster et al 1319

5 (42.1%) of 57 mesotheliomas, despite their attempts to optimize immunostaining. Our work with the Zymed anticalretinin antibody showed stronger staining in some cases, but no dramatic difference between the Zymed and the Chemicon anticalretinin antibodies. The conflicting reports surrounding calretinin, in addition to our experience, suggest that the particular anticalretinin antibody employed, as well as perhaps the antigen retrieval method used, may influence the results. 29 This study supports the prior conclusions by Brown et al 35 regarding the lack of utility of thrombomodulin and indeed reinforces their assertion that a single immunohistochemical marker, which is positive in all mesotheliomas and negative in all adenocarcinomas may not exist. However, nuclear WT1 expression has good sensitivity and specificity for the diagnosis of malignant mesotheliomas. While this is not sufficient to stand alone as a differentiating marker in all cases, in the appropriate clinical setting, our findings show that positive nuclear staining is strongly confirmatory of malignant mesothelioma. References 1. Morgan RL, DeYoung BR, McGaughy VR, Niemann TH. Moc-31 aids in the differential between adenocarcinoma and reactive mesothelial cells. Cancer. 1999;87: Ordoñez, Nelson G. Value of calretinin immunostaining in differentiating epithelial mesothelioma from lung adenocarcinoma. Mod Pathol. 1998;11: Ordoñez, Nelson G. Value of Ber-EP4 antibody in differentiating epithelial pleural mesothelioma from adenocarcinoma. Am J Clin Pathol. 1998;109: Ordoñez, Nelson G. Value of thrombomodulin immunostaining in the diagnosis of mesothelioma. Histopathology. 1997;31: Doglioni C, Dei Tos P, Laurino L, et al. Calretinin: a novel immunocytochemical marker for mesothelioma. Am J Surg Pathol. 1996;20: Fetsch PA, Abati A, Hijazi YM. Utility of the antibodies CA 19 9, HBME-1, and thrombomodulin in the diagnosis of malignant mesothelioma and adenocarcinoma in cytology. Cancer. 1998;84: Garcia-Prats MD, Ballestin C, Sotelo T, Lopez-Encuentra A, Mayordomo JI. A comparative evaluation of immunohistochemical markers for the differential diagnosis of malignant pleural tumors. Histopathology. 1998;32: Gotzos V, Vogt P, Celio MR. The calcium binding protein calretinin is a selective marker for malignant pleural mesotheliomas of the epithelial type. Pathol Res Pract. 1996;192: Attanoos RL. -binding antibodies: thrombomodulin, OV-632, and HBME-1 and their use in the diagnosis of malignant mesothelioma. Histopathology. 1996;29: Kennedy AD, King G, Kerr K. HBME-1 and antithrombomodulin in the differential diagnosis of malignant mesothelioma of the pleura. Am J Clin Pathol. 1997;50: Attanoos RL, Webb R, Gibbs AR. CD44H expression in reactive mesothelium, pleural mesothelioma and pulmonary adenocarcinoma. Histopathology. 1997;30: Ordoñez, Nelson G. Value of the MOC-31 monoclonal antibody in differentiating epithelial pleural mesothelioma from lung adenocarcinoma. Hum Pathol. 1998;29: Ordoñez, Nelson G. Value of thyroid transcription factor-1, E-cadherin, BG8, WT-1, and CD44S immunostaining in distinguishing epithelial pleural mesothelioma from pulmonary and nonpulmonary adenocarcinoma. Am J Surg Pathol. 2000;24: Brockstedt U, Gulyas M, Dobra K, Dejmek A, Hjerpe A. An optimized battery of eight antibodies that can distinguish most cases of epithelial mesothelioma from adenocarcinoma. Am J Clin Pathol. 2000;114: Carella R, Delonardi G, D Errico A, et al. Immunohistochemical panels for differentiating epithelial malignant mesothelioma from lung adenocarcinoma. Am J Surg Pathol. 2001;25: Moran CA, Wick MR, Suster S. The role of immunohistochemistry in the diagnosis of malignant mesothelioma. Semin Diagn Pathol. 2000;17: Pritchard-Jones K, Fleming S, Davidson D, et al. The candidate Wilms tumor gene is involved in genitourinary development. Nature. 1990;345: Huang A, Campbell CE, Bonetta L, et al. Tissue, developmental and tumorspecific expression of divergent transcripts in Wilms tumor. Science. 1990;250: Park S, Schalling M, Bernard A, et al. The Wilms tumor gene WT1 is expressed in murine mesoderm-derived tissues and mutated in a human mesothelioma. Nat Genet. 1993;4: Eccles MR, Yun K, Reeve AE, Fidler AE. Comparative in situ hybridization analysis of PAX2, PAX8, and WT1 gene transcription in human fetal kidney and Wilms tumors. Am J Pathol. 1995;146: Walker C, Rutten F, Yuan X, Puss H, Meev DM, Everitt J. Wilms tumor suppressor gene expression in rat and human mesothelioma. Cancer Res. 1994; 54: Langerak AW, Williamson KA, Miyagawa K, Hagemeijer A, Versned MA, Hastie ND. Expression of the Wilms tumor gene WT1 in human malignant mesothelioma cell lines and relationship to platelet-derived growth factor A and insulin-like growth factor 2 expression. Genes Chromosom Cancer. 1995;12: Pearson GR, Gregory SP, Charles AK. Immunohistochemical demonstration of Wilms tumor gene product WT1 in a canine neuroepithelioma: providing evidence for its classification as an extrarenal nephroblastoma. J Comp Pathol. 1997;116: Oji Y, Ogawa H, Tamaki H, et al. Expression of the Wilms tumor gene WT-1 in solid tumors and its involvement in tumor cell growth. Jpn J Cancer Res. 1999;90: Kumar-Singh S, Segers K, Rodeck U, et al. WT1 mutation in malignant mesothelioma and WT1 immunoreactivity in relation to p53 and growth factor receptor expression cell-type transition, and prognosis. J Pathol. 1997;181: Larsson SH, Charlieu JP, Miyagawa K, et al. Subnuclear localization of WT1 in splicing or transcription factor domains is regulated by alternative splicing. Cell. 1995;81: Amin KM, Litzky LA, Smythe WR, et al. Wilms tumor 1 susceptibility (WT1) gene products are selectively expressed in malignant mesothelioma. Am J Pathol. 1995;146: Oates J, Edwards C. HBME, MOC-31 and calretinin: an assessment of recently described markers for mesothelioma and adenocarcinoma. Histopathology. 2000;36: Ordoñez, Nelson G. In search of a positive immunohistochemical marker for mesothelioma: an update. Adv Anat Pathol. 1998;5: Ordoñez, Nelson G. The immunohistochemical diagnosis of epithelial mesothelioma. Hum Pathol. 1999;30: Barberis MCP, Faleri M, Veronese S, et al. Calretinin: a selective marker of normal and neoplastic mesothelial cells in serous effusions. Acta Cytol. 1997;41: Leers MPG, Aarts MMJ, Thuenissen PHMH, et al. E-cadherin and calretinin: a useful combination of immunochemical markers for differentiation between mesothelioma and metastatic adenocarcinoma. Histopathology. 1998;32: Cury PM, Butcher DN, Fisher C, Corring B, Nicholson AG. Value of the mesothelium-associated antibodies thrombomodulin, cytokeratin 5/6, calretinin, and CD44H in distinguishing epithelioid pleural mesothelioma from adenocarcinoma metastatic to the pleura. Mod Pathol. 2000;13: Riera JR, Atengo-Osuna C, Longmate JA, Battifora H. The immunohistochemical diagnostic panel for epithelial mesothelioma: a reevaluation after heatinduced epitope retrieval. Am J Surg Pathol. 1997;21: Brown RW, Clark GM, Tandon AK, Allred DC. Multiple-marker immunohistochemical phenotypes distinguishing malignant pleural mesothelioma from pulmonary adenocarcinoma. Hum Pathol. 1993;24: Arch Pathol Lab Med Vol 125, October 2001 Immunohistochemical Analysis of WT1 Staining Foster et al