Hepatocyte Nuclear Factor-1b Is Not a Specific Marker of Clear Cell Carcinoma in Serous Effusions

Transcription

1 Hepatocyte Nuclear Factor-1b Is Not a Specific Marker of Clear Cell Carcinoma in Serous Effusions Ben Davidson, MD, PhD 1,2 BACKGROUND: The transcription factor hepatocyte nuclear factor-1b (HNF1b) has been reported to be a specific clear cell carcinoma marker, but its role in the diagnosis of serous effusions is largely unexplored. The objective of this study was to assess the diagnostic role of 2 commercial antibodies against HNF1b in effusion specimens. METHODS: Effusions (n 5 101), consisting of 43 ovarian adenocarcinomas (26 serous, 14 clear cell, 3 endometrioid), 37 nonovarian adenocarcinomas, 10 malignant mesotheliomas, 2 nonepithelial cancers, and 9 reactive specimens, were immunostained using antibodies from Santa Cruz Biotechnology Inc (Santa Cruz, Calif) and Atlas Antibodies AB (Stockholm, Sweden). RESULTS: Use of the Santa Cruz antibody was associated with cytoplasmic or background staining in some specimens, whereas distinct staining with minimal background was observed using the Atlas antibody. The Santa Cruz antibody performed better in differentiating clear cell carcinoma from serous ovarian carcinoma and breast carcinoma, whereas staining was consistently negative in benign and malignant mesotheliomas using both antibodies. Distinct nuclear expression of HNF1b was observed in lung and gastrointestinal carcinomas, most often using the Atlas antibody. CONCLUSIONS: The HNF1b antibody from Atlas performed better than its counterpart from Santa Cruz in terms of staining quality, but was less specific for clear cell carcinoma. Although HNF1b may be of diagnostic value in differentiating clear cell from serous carcinoma in cases with proven genital origin, the role of this marker is questionable in the differential diagnosis between the former tumors and adenocarcinomas of other origin, particularly in the setting of metastasis from an unknown primary tumor. Cancer (Cancer Cytopathol) 2014;122: VC 2013 American Cancer Society. KEY WORDS: hepatocyte nuclear factor-1b; effusions; adenocarcinoma; clear cell; diagnosis. INTRODUCTION The serosal cavities are frequently affected by cancer, including both primary and metastatic tumors. Metastases, which by far outnumber primary cancers, are predominantly from carcinomas, of which the most frequent ones are lung, breast, ovary, or gastrointestinal (GI) origin. Serous effusions are not infrequently the initial specimen on which the diagnosis of malignancy is made, as well as frequent site of metastasis and recurrence for previously diagnosed cancers. The diagnosis of serous effusions is currently based mainly on morphological evaluation and immunohistochemistry (IHC), and IHC panels have expanded in recent years to include an array of originspecific markers. 1 Marker specificity, however, depends, in addition to technical factors, on the tumors compared. Hepatocyte nuclear factor-1b (HNF1b) is a transcription factor expressed in the liver and kidney during embryogenesis. In adult tissues, HNF1b is detected in epithelial cells of the liver and kidney, in other Corresponding author: Ben Davidson, MD, PhD, Division of Pathology, Norwegian Radium Hospital, Oslo University Hospital, Montebello N-0310 Oslo, Norway; Fax: 011 (47) ; bend@medisin.uio.no 1 Division of Pathology, Oslo University Hospital, Norwegian Radium Hospital, Oslo, Norway; 2 University of Oslo, Faculty of Medicine, Institute of Clinical Medicine, Oslo, Norway The competent work of Ms. Helene Tuft Stavnes and Mr. Arild Holth in performing the immunohistochemistry analysis is gratefully acknowledged. Received: July 18, 2013; Revised: August 10, 2013; Accepted: August 14, 2013 Published online September 24, 2013 in Wiley Online Library (wileyonlinelibrary.com) DOI: /cncy.21353, wileyonlinelibrary.com Cancer Cytopathology February

2 TABLE 1. Case Distribution According to Tumor Origin and Effusion Site Site Clinical Diagnosis Histology No. Peritoneum Pleura Pericardium Ovarian carcinoma Serous Clear Endometrioid Endometrial carcinoma AC a Uterine cervical carcinoma AC Breast carcinoma AC b Lung carcinoma AC Gastrointestinal carcinomas c AC Malignant mesothelioma Epithelioid Other cancers d Reactive Total a Adenocarcinoma (AC), including 4 clear cell carcinomas, 1 serous carcinoma, and 1 poorly differentiated carcinoma, not otherwise specified. b Including 8 ductal carcinomas, 1 lobular carcinoma, and 1 carcinoma of mixed type. c Including 5 esophageal, 2 colon, 2 gastric, and 1 pancreatic carcinoma. d Including 1 prostate carcinoma, 1 malignant melanoma, and 1 granulosa cell tumor. genitourinary and GI organs, and in the endometrium. Genes regulated by HNF1b include those coding for dipeptidyl peptidase IV (also known as CD26), osteopontin, annexin A4, and angiotensin converting enzyme 2. Cellular functions attributed to HNF1b include regulation of proliferation, apoptosis, and glucose homeostasis; preservation of the epithelial phenotype; and response to oxidative stress. 2,3 HNF1b is expressed in endometriosis, 4 as well as in clear cell carcinomas (CCC) of the ovary, 4-9 endometrium, 6,10 uterine cervix, 11 kidney, 9,12 and urinary bladder and urethra, 13 as well as in hepatocellular carcinomas with clear cytoplasm. 9 Although several of the aforementioned studies reported HNF1b to be a specific CCC marker, 2 studies have documented staining in tumors of other histology in the uterine cervix and endometrium. 10,11 The diagnostic role of HNF1b in effusion cytology has, to date, been assessed in a single study, in which 60 malignant ascites specimens from ovarian carcinoma patients were immunostained. In the study, by Higashiguchi et al, HNF1b expression was reported to be specific for CCC, with the exception of 3 endometrioid carcinomas. 14 The study, however, did not analyze the role of this marker in distinguishing ovarian from nonovarian cancers. In the present study, the diagnostic role of HNF1b using 2 commercial antibodies was evaluated in 101 effusion specimens, consisting of both ovarian and nonovarian malignancies, as well as benign reactive specimens. The data presented suggest that whereas HNF1b may differentiate CCC from serous ovarian carcinoma and cells of mesothelial origin, it performs less well in the differential diagnosis with carcinomas of other origin. MATERIALS AND METHODS Effusion Specimens Effusions (n 5 101) were submitted to the Department of Pathology at the Norwegian Radium Hospital, Oslo, Norway, for routine diagnostic purposes during 1998 through Specimens were centrifuged and cell blocks were prepared using the thrombin clot method. Diagnoses were established by morphological evaluation of smears and cell block sections from formalin-fixed paraffin-embedded pellets and by IHC, and were confirmed in surgical specimens. Diagnoses and specimen site are detailed in Table 1. The Regional Committee for Medical Research Ethics in Norway approved the study. Immunohistochemistry Paraffin-embedded sections from 101 effusions were immunostained with 2 commercial antibodies against HNF1b, a goat polyclonal antibody (catalog no. sc-7411, C-20) from Santa Cruz Biotechnology (Santa Cruz, Calif), diluted 1:2000, and an affinity-isolated rabbit antibody (catalog no. HPA002083) from Atlas Antibodies (Stockholm, Sweden), diluted 1:300. Sections were pretreated with high ph (Santa Cruz antibody) or low ph (Atlas antibody) buffer in the Dako PT-Link (Dako, Glostrup, Denmark) and incubated with 154 Cancer Cytopathology February 2014

3 HNF1b in Effusions/Davidson TABLE 2. Hepatocyte Nuclear Factor-1b Immunostaining Results a Atlas Antibodies Santa Cruz Biotechnology Clinical Diagnosis No Ovarian carcinoma, serous Ovarian carcinoma, clear cell Ovarian carcinoma, endometrioid Endometrial carcinoma Uterine cervical carcinoma Breast carcinoma Lung carcinoma Gastrointestinal carcinomas b Malignant mesothelioma Other cancers c Reactive Total a Staining score: 0 5 no staining, 1 5 1%-10% of cells, %-100% of cells. b Including 5 esophageal, 2 colon, 2 gastric, and 1 pancreatic carcinoma. c Including 2 prostate carcinomas, 1 malignant melanoma, and 1 granulosa cell tumor. 0.03% hydrogen peroxide to inhibit endogenous peroxidase. Sections were subsequently incubated with the primary antibody for 30 minutes. Visualization was achieved using the Dako EnVision FLEX1 peroxidase system. Positive controls consisted of an ovarian CCC previously shown to express HNF1b. Negative controls consisted of sections that underwent similar staining procedure with nonrelevant rabbit or goat immunoglobulins. Staining was optimized by testing multiple dilutions of both antibodies to achieve discernible nuclear staining with minimal background. Staining was considered positive only when unequivocally localized to the cell nucleus. Staining extent was scored as negative (0% of cells), focal (1%-10% of cells), or diffuse (> 10% of cells). No specimen contained fewer than 100 tumor cells. RESULTS HNF1b immunostaining results are detailed in Table 2 and illustrated in Figure 1. Nuclear staining was seen in 12 of 14 ovarian CCC using both antibodies, with the majority (often 100%) of cells staining positive (Fig. 1A,B). The 2 negative CCC were the same using both antibodies. The 3 endometrioid carcinomas showed variable staining (Fig. 1E,F). Notable difference between the 2 antibodies was seen with respect to serous ovarian carcinomas, with staining of 5 of 26 versus 0 of 26 using the Atlas and Santa Cruz antibodies, respectively. Another feature common to both antibodies was uniformly negative staining in reactive mesothelial cells and in malignant mesotheliomas. However, the Santa Cruz antibody stained leukocytes in some specimens and generated background staining in others (Fig. 1N,O). Both antibodies stained all 4 uterine CCC (Fig. 1C,D), as well as the serous endometrial carcinoma, whereas the poorly differentiated carcinoma case was HNF1b-negative using both antibodies. Among cervical carcinomas, 2 specimens from the same patient were strongly HNF1b-positive using both antibodies (Fig. 1P). This patient has an adenocarcinoma of endocervical type (Fig. 1Q), but later developed a large (12 cm) tumor in the ovary, which was diagnosed as metastasis from the cervical tumor. Upon reevaluation of this specimen by the author, the morphology of these tumors was interpreted as different, with the ovarian tumor containing areas with features of CCC with eosinophilic cells (Fig. 1R). Because the effusions were tapped after the diagnosis of the ovarian tumor, it is possible that the latter was the true origin of the HNF1b-positive tumor cells. Evaluation of the adenocarcinomas of GI, lung, and breast origin showed frequent HNF1b expression applying the Atlas antibody, which was most pronounced in GI carcinomas, regardless of their site of origin. Staining in these specimens was generally seen in > 50% of cells. The Santa Cruz antibody did not stain any of the 10 breast carcinomas, but was immunoreactive with a substantial number of lung and GI carcinomas, as well as in a case of metastatic melanoma (Fig. 1G-M). DISCUSSION Establishing the origin of a newly diagnosed metastatic cancer continues to be a challenging task in some cases, Cancer Cytopathology February

4 FIGURE 1. Micrographs show HNF1b expression in serous effusions. (A,B) There is strong nuclear staining in tumor cells in ovarian clear cell carcinoma using the antibodies from (A) Atlas Antibodies and (B) Santa Cruz Biotechnology, and (C,D) a similar staining pattern appears in an endometrial clear cell carcinoma using the antibodies from (C) Atlas Antibodies and (D) Santa Cruz Biotechnology. (E,F) Ovarian endometrioid carcinoma is shown expressing HNF1b using the (E) Atlas and (F) Santa Cruz antibodies. (G,H) In breast carcinoma, images show (G) a case of metastatic ductal carcinoma expressing HNF1b using the Atlas antibody. (H) The same tumor was HNF1b-negative using the Santa Cruz antibody. (I-L) HNF1b-positive nuclei are shown using the Atlas antibody in (I) esophageal, (J) lung, (K) gastric, and (L) colon adenocarcinoma. (M) HNF1b-positive lung adenocarcinoma using the antibody from Santa Cruz Biotechnology. (N) Background staining is shown in a uterine adenocarcinoma effusion using the Santa Cruz antibody. Tumor cells stain strongly at the nucleus. (O) Reactive effusion was immunostained for HNF1b using the Santa Cruz antibody. Lymphocytes stain strongly positive at the cytoplasm. (P-R) Micrographs are from cervical carcinoma. HNF1b expression (Atlas antibody) is shown in an effusion specimen diagnosed as metastatic cervical carcinoma (P). The tumor in the uterine cervix was an adenocarcinoma of endocervical type (Q). However, the patient later presented with a 12-cm large ovarian tumor with areas raising suspicion of a clear cell carcinoma (R). and many markers initially reported to be highly specific for a given entity have later been shown to stain multiple other tumors. Two important markers, WT1 (Wilms tumor 1) and PAX8 (paired box 8), have nevertheless been found to be of value in classifying carcinomas of the female genital system. WT1 is a sensitive and specific marker of serous carcinoma, whereas PAX8 is a general marker of tumors of female genital origin. Although both markers may stain other cancers (eg, PAX8 in renal cell carcinoma), they are highly useful in effusion cytology, when applied within the relevant clinical context A large number of studies have shown HNF1b to be a sensitive CCC marker, 4-14 albeit with no ability to identify the organ of origin of such tumors. The majority of these studies have additionally postulated HNF1b to be a specific marker of CCC, although this has been challenged in the case of endometrial and cervical carcinoma. 10,11 The antibody used in previous articles was the Santa Cruz antibody used in the present study. This study is the first to compare the performance of 2 commercial HNF1b antibodies, as well as to analyze the expression of this protein in nongenital cancers in serous effusions. Results demonstrate that both antibodies have limitations that question their applicability to the differential diagnosis of serous effusions. The HNF1b antibody by Atlas was highly specific in identifying tumor cells as adenocarcinoma, because malignant mesotheliomas and reactive mesothelial cells 156 Cancer Cytopathology February 2014

5 HNF1b in Effusions/Davidson FIGURE 1. (Continued) were consistently negative. However, the differentiation between CCC and cells of mesothelial origin is rarely difficult even based on morphology alone. In the main role attributed to HNF1b, namely the differentiation between CCC and serous carcinoma, the performance of this antibody was moderately good at best, in view of the fact that Cancer Cytopathology February

6 5 of 26 serous ovarian carcinomas stained strongly and diffusely. In addition, the frequent staining of adenocarcinomas of breast, lung, and GI origin puts into serious question the role of this marker in identifying the primary organ of malignant cells of uncertain origin. The HNF1b antibody by Santa Cruz performed well in the differential diagnosis from serous ovarian carcinoma, malignant mesothelioma, reactive mesothelium, and breast carcinoma, with 100% specificity in differentiating these entities from CCC. However, it performed less well in the differential diagnosis from lung and GI tract adenocarcinomas. In addition, it was associated with considerable background staining in some specimens, even at optimized dilution. The suboptimal staining quality using this antibody may prove problematic when considering its inclusion in diagnostic panels. In conclusion, the diagnostic role of HNF1b in serous effusions was analyzed using 2 commercial antibodies. Results suggest that the main role of this marker in effusion diagnosis is the differentiation between CCC and serous carcinoma, preferably in combination with a serous carcinoma marker, such as WT1. HNF1b is additionally of value in differentiating CCC from benign or malignant mesothelial cells, although this diagnostic difficulty is infrequently encountered. The Santa Cruz antibody additionally appears to differentiate CCC from breast carcinoma. However, in view of the generally frequent expression of HNF1b in carcinomas of nongenital origin, there appears to be a limited value for this marker in broader differential diagnostic settings. Future studies are needed to evaluate the sensitivity and specificity of this marker. FUNDING SOURCES This work was supported by the Inger and John Fredriksen Foundation for Ovarian Cancer Research. CONFLICT OF INTEREST DISCLOSURE The author made no disclosures. REFERENCES 1. Davidson B, Firat P, Michael CW, eds. Serous Effusions. London, UK: Springer; Kobayashi H, Yamada Y, Kanayama S, et al. The role of hepatocyte nuclear factor-1beta in the pathogenesis of clear cell carcinoma of the ovary. Int J Gynecol Cancer. 2009;19: Kajihara H, Yamada Y, Kanayama S, et al. Clear cell carcinoma of the ovary: potential pathogenic mechanisms (Review). Oncol Rep. 2010;23: Kato N, Sasou S, Motoyama T. Expression of hepatocyte nuclear factor-1beta (HNF-1beta) in clear cell tumors and endometriosis of the ovary. Mod Pathol. 2006;19: Tsuchiya A, Sakamoto M, Yasuda J, et al. Expression profiling in ovarian clear cell carcinoma: identification of hepatocyte nuclear factor-1 beta as a molecular marker and a possible molecular target for therapy of ovarian clear cell carcinoma. Am J Pathol. 2003;163: Yamamoto S, Tsuda H, Aida S, Shimazaki H, Tamai S, Matsubara O. Immunohistochemical detection of hepatocyte nuclear factor 1beta in ovarian and endometrial clear-cell adenocarcinomas and nonneoplastic endometrium. Hum Pathol. 2007;38: K obel M, Kalloger SE, Carrick J, et al. A limited panel of immunomarkers can reliably distinguish between clear cell and highgrade serous carcinoma of the ovary. Am J Surg Pathol. 2009;33: DeLair D, Oliva E, K obel M, Macias A, Gilks CB, Soslow RA. Morphologic spectrum of immunohistochemically characterized clear cell carcinoma of the ovary: a study of 155 cases. Am J Surg Pathol. 2011;35: Kao YC, Lin MC, Lin WC, Jeng YM, Mao TL. Utility of hepatocyte nuclear factor-1b as a diagnostic marker in ovarian carcinomas with clear cells. Histopathology. 2012;61: Fadare O, Liang SX. Diagnostic utility of hepatocyte nuclear factor 1-Beta immunoreactivity in endometrial carcinomas: lack of specificity for endometrial clear cell carcinoma. Appl Immunohistochem Mol Morphol. 2012;20: Park KJ, Kiyokawa T, Soslow RA, et al. Unusual endocervical adenocarcinomas: an immunohistochemical analysis with molecular detection of human papillomavirus. Am J Surg Pathol. 2011;35: Sangoi AR, Fujiwara M, West RB, et al. Immunohistochemical distinction of primary adrenal cortical lesions from metastatic clear cell renal cell carcinoma: a study of 248 cases. Am J Surg Pathol. 2011;35: Brimo F, Herawi M, Sharma R, Netto GJ, Epstein JI, Illei PB. Hepatocyte nuclear factor-1b expression in clear cell adenocarcinomas of the bladder and urethra: diagnostic utility and implications for histogenesis. Hum Pathol. 2011;42: Higashiguchi A, Yamada T, Susumu N, et al. Specific expression of hepatocyte nuclear factor-1beta in the ovarian clear cell adenocarcinoma and its application to cytological diagnosis. Cancer Sci. 2007;98: Han L, Pansare V, Al-Abbadi M, Husain M, Feng J. Combination of MUC5ac and WT-1 immunohistochemistry is useful in distinguishing pancreatic ductal carcinoma from ovarian serous carcinoma in effusion cytology. Diagn Cytopathol. 2010;38: McKnight R, Cohen C, Siddiqui MT. Utility of paired box gene 8 (PAX8) expression in fluid and fine-needle aspiration cytology: an immunohistochemical study of metastatic ovarian serous carcinoma. Cancer Cytopathol. 2010;118: Tong GX, Devaraj K, Hamele-Bena D, et al. PAX8: A marker for carcinoma of M ullerian origin in serous effusions. Diagn Cytopathol. 2011;39: Wiseman W, Michael CW, Roh MH. Diagnostic utility of PAX8 and PAX2 immunohistochemistry in the identification of metastatic M ullerian carcinoma in effusions. Diagn Cytopathol. 2011;39: Cancer Cytopathology February 2014