- Email: [email protected]
Expression of the Ets transcription factor EHF in serous ovarian carcinoma effusions is a marker of poor survival
Human Pathology (2012) 43, 496–505
www.elsevier.com/locate/humpath
Original contribution
Expression of the Ets transcription factor EHF in serous ovarian carcinoma effusions is a marker of poor survival☆ Kjersti Brenne MSc a , Dag André Nymoen MSc a , Thea Eline Hetland MD b , Claes G. Trope' MD, PhD b,c , Ben Davidson MD, PhD a,c,⁎ a
Division of Pathology, Norwegian Radium Hospital, Oslo University Hospital, N-0424 Oslo, Norway Department of Gynecologic Oncology, Norwegian Radium Hospital, Oslo University Hospital, N-0424 Oslo, Norway c University of Oslo, Faculty of Medicine, Institute of Clinical Medicine, N-0424 Oslo, Norway b
Received 22 January 2011; revised 19 May 2011; accepted 20 May 2011
Keywords: Ovarian carcinoma; Malignant mesothelioma; Diagnosis; Survival; Quantitative real-time PCR
Summary The EHF (Ets homologous factor) gene was previously shown to be overexpressed in ovarian/ primary peritoneal serous carcinoma compared to malignant mesothelioma using gene expression arrays. The objective of this study was to validate this finding at the mRNA level in a larger series. We analyzed the diagnostic role of EHF in 98 ovarian serous carcinoma effusions, 23 malignant mesothelioma specimens (20 effusions, 3 surgical specimens), and 28 primary ovarian serous carcinomas using quantitative real-time polymerase chain reaction. Expression levels of EHF in ovarian carcinoma were additionally investigated for association with clinicopathologic parameters and survival. Quantitative real-time polymerase chain reaction analysis showed significantly higher expression of EHF mRNA in ovarian carcinoma effusions and in primary ovarian carcinoma compared to malignant mesothelioma effusions (P b .001 for both). EHF mRNA expression was additionally higher in primary ovarian carcinomas compared to effusions of this cancer (P b .001). In univariate analysis for all patients with effusions, higher EHF mRNA levels were associated with a trend for shorter progression-free survival (P = .066), which became significant in analysis of 45 patients with primary diagnosis pre-chemotherapy effusions (P = .01). In Cox multivariate analysis, EHF mRNA expression was an independent predictor of poor progression-free survival for all patients and patients with primary diagnosis pre-chemotherapy effusions (P = .033 and P = .009, respectively). EHF mRNA levels differentiate ovarian carcinoma from malignant mesothelioma and may thus be of diagnostic value in this setting. EHF may be a novel prognostic marker in ovarian carcinoma. © 2012 Elsevier Inc. All rights reserved.
Abbreviations: EHF, Ets homology factor; qPCR, quantitative real-time PCR; EBS, Ets binding site; ESE-3, epithelial specific ETS-3; OC, ovarian carcinoma; PPC, primary peritoneal carcinoma; MM, malignant mesothelioma; FIGO, International Federation of Gynecology and Obstetrics; H&E, hematoxylin and eosin; PFS, progression-free survival; OS, overall survival. ☆ This work was supported by grants from the Norwegian Cancer Society and the Research Fund at the Norwegian Radium Hospital, and by the Inger and John Fredriksen Foundation for Ovarian Cancer Research. ⁎ Corresponding author. Department of Pathology, Norwegian Radium Hospital, Oslo University Hospital, Montebello N-0310 Oslo, Norway. E-mail address: [email protected] (B. Davidson). 0046-8177/$ – see front matter © 2012 Elsevier Inc. All rights reserved. doi:10.1016/j.humpath.2011.05.023
EHF in serous ovarian carcinoma
1. Introduction Ets homologous factor (EHF) is member of the ETS gene family, one of the largest families of transcriptional regulators. The ETS gene family was discovered through the identification of the viral oncogene v-ets in the avian transforming retrovirus E26, and is divided into subfamilies (ELF, ELG, ERG, ERF, ESE, ETS, PEA3, ER71, SPI, TCF, TEL) based on sequence similarity in the ETS domain. The ETS domain is an evolutionarily conserved winged helix-turn-helix DNA binding domain consisting of 85 amino acids that is present in all ETS proteins. This DNA binding domain recognizes a core GGAA/T sequence known as the Ets binding site (EBS) [1-7]. All ETS transcription factors bind to the EBS, and such EBS have been identified in the promoter and enhancer regions of viral and cellular genes [5]. ETS gene products activate or repress expression of genes that are involved in cellular proliferation, differentiation, hematopoiesis, apoptosis, tissue remodeling, angiogenesis, transformation, invasion and metastasis. EHF, also known as ESE-3 (epithelial specific ETS-3), is a member of the ESE subfamily which is epithelial-specific and has distinct functions in epithelial cell differentiation and proliferation [8]. As a result of its involvement in these cellular processes, uncoordinated expression of EHF has the potential to contribute to tumor growth and consequently disease progression. Ovarian carcinoma (OC) is the second most common gynecologic cancer, and has the highest death toll among gynecologic cancers, with more than 50% of diagnosed women dying of their disease [9]. The poor outcome of OC reflects the fact that most patients present with advanced stage (III/IV) disease. Despite efforts to improve disease outcome by focus on early detection and the combination of aggressive surgery with improved chemotherapy protocols, the mortality rate of OC has remained largely unchanged. The serosal cavities, including the peritoneal, pleural, and pericardial spaces, are frequently affected by cancer. The presence of cancer cells in effusions at these anatomic sites is evidence of advanced-stage disease with metastatic spread, and regardless of the tumor site of origin, marks disease progression and is associated with poor survival [10,11]. One of the most challenging differential diagnoses in serosal cancer is between ovarian/primary peritoneal serous carcinoma (OC/PPC) and malignant mesothelioma (MM), tumors that are closely related in terms of clinical presentation, morphology, and immunohistochemical phenotype. We recently reported on the differential expression of 189 genes in OC/PPC compared with diffuse peritoneal MM using cDNA microarray technology [12]. Among the differentially expressed genes, EHF was identified as a gene that is significantly overexpressed in OC/PPC effusions compared to peritoneal MM effusions. The present study evaluated the diagnostic value of EHF mRNA expression
497 levels in differentiating OC from MM and analyzed the prognostic value of EHF in OC.
2. Materials and methods 2.1. Patients and material 2.1.1. Effusions The material analyzed in the present study consisted of 112 fresh-frozen effusions, including 98 Müllerian carcinomas and 23 MMs, submitted for routine diagnostic purposes to the Division of Pathology, Norwegian Radium Hospital, during the period 1998 to 2005. Müllerian carcinoma effusions (74 peritoneal, 24 pleural) were obtained from 93 patients (5 patients with 2 effusions each) diagnosed with International Federation of Gynecology and Obstetrics (FIGO) stage II-IV OC (n = 77), PPC (n = 11), or carcinoma of the fallopian tube (n = 5). Most of the specimens (82/98; 84%) were of the serous type (Fig. 1A-C). Because of their closely linked histogenesis and phenotype, all these tumors are referred to as OC in the following sections. Clinicopathologic data for patients with OC effusions were obtained from the Department of Gynecologic Oncology at the Norwegian Radium Hospital and are summarized in Table 1. Most of the patients (83/93; 89%) received platinum-based chemotherapy at diagnosis. The 23 MM specimens consisted of 20 effusions (15 pleural, 5 peritoneal) (Fig. 1D-F) and 3 surgical specimens of diffuse peritoneal MM. All were from patients diagnosed with MM of the epithelioid or biphasic type in biopsy specimens. Effusion specimens were centrifuged immediately after tapping, and cell pellets were frozen at −70°C in equal amounts of RPMI 1640 medium containing 50% fetal calf serum and 20% dimethylsulfoxide. Smears and hematoxylin and eosin (H&E)–stained cell block sections were reviewed by a surgical pathologist experienced in cytopathology (B. D.). Diagnoses were based on morphology and immunohistochemistry. The antibody panel used for differentiating OC/PPC from MM and for establishing the malignant character of the mesothelial cell proliferations included Ber-EP4, B72.3, calretinin, desmin, and EMA. BG8 was added as a carcinoma marker in some of the cases, whereas WT-1 was used as marker of serous differentiation in OC specimens. The tumor cell population was more than 50% in all specimens analyzed in this study. 2.1.2. Primary OC In addition to the above-mentioned effusions, 28 solid primary OC, all of the serous type, were examined (Fig. 1G-I). Histologic grade was as follows: grade 1, five tumors; grade 2, five tumors; grade 3, eighteen tumors. FIGO stage was as follows: I, 1 patient; II, 1 patient; III, 20 patients; IV, 6 patients. Residual disease volume was available for 24 patients, of whom 16 were debulked to ≤1
498
K. Brenne et al.
Fig. 1 Morphology. Examples of 3 serous carcinoma effusions (A-C), 3 MM effusions (D-F), and 3 primary ovarian serous carcinomas (G-I); A and D, PAP stain; B and E, Diff-Quik stain; C, F to I, H&E.
cm and 8 to N1 cm. Surgical specimens were snap-frozen and kept at −70°C with no medium. Frozen sections from all tumors were evaluated for the presence of a N50% tumor component and absence of necrosis. H&E-stained sections from these tumors were reviewed to establish tumor type and histologic grade. The same procedure was applied to the 3 solid MM specimens.
2.2. Quantitative real-time PCR Total RNA was extracted using the automated sample preparation instrument QiaCube and the RNeasy Mini Kit
(Qiagen, Hilden, Germany). mRNA was reverse-transcribed into cDNA using Superscript III Reverse Transcriptase (Invitrogen, Carlsbad, CA). The EHF (NM_012153.3) assay covered exon junction 12. Primer specificity and efficiency were validated by gel electrophoresis and Power SYBR Green (Applied Biosystems, Foster City, CA), respectively. The assay was controlled for primer dimers using the NetPrimer software by PREMIER Biosoft, as well as for single nucleotide polymorphisms through the NCBI database. Primer efficiency was tested using Power SYBR Green (Applied Biosystems) and a dilution series of synthetic oligonucleotide as template and subsequently as standard curve. The quantitative real-time
EHF in serous ovarian carcinoma
499
Table 1 Clinicopathologic data of the OC effusion cohort (n = 93 patients) Parameter
No. of patients
Age
34-87 y (mean = 61) 1 53 39 7 20 56 10 78 11 4 32 47 14 49 40 4
FIGO stage
Grade
Histology
Residual disease
Chemoresponse at diagnosis
II III IV I II III NA a Serous Non-serous b NA c ≤1 cm N1 cm NA d Complete Incomplete e ND f
a NA = nonavailable; including effusions from inoperable patients where biopsy was too small for grading and patients operated in other hospitals, for which the primary tumor could not be accessed for assessment of grade. b Including clear cell, endometrioid, undifferentiated, and mixed histology carcinomas. c Patients with histologically diagnosed OC for which the primary tumor could not be accessed for assessment of tumor type. d Patients who were inoperable or had no record. e Partial response/stable disease/progression/allergic or adverse reaction. f ND = nondetermined, including patients who received no chemotherapy and patients who died before chemoresponse could be assessed.
PCR (qPCR) reaction was run using the Platinum qPCR SuperMix-UDG with ROX solution (Invitrogen) and quantified on the Applied Biosystems 7900HT Sequence Detection System. Primer and probe sequences as follows: Forward primer 5′- ACCCAGAATCTTTAGGTAAATGAGATC- 3′ Reverse primer 5′- CTGGTGAAGGAGGTTGTTGC-3′ Probe Fam 5′- ATTCTGGAAGGAGGTGGTGTAATGAATCTCAACC-3′ nonfluorescent quencher
An array of 12 reference genes (TaqMan low-density array human endogenous control panel; Applied Biosystems) was tested to identify the most uniformly expressed transcript in effusions specimens. Based on these results, the expression levels of EHF were normalized against the housekeeping gene β-glucuronidase (GUS). The GUS primer and probe sequences have been published elsewhere [13]. Standard curves for the GUS assay were purchased from Ipsogen (Marseille, France).
Synthetic oligonucleotides were used as standard curve for EHF. Standard curves were used to obtain copy number of EHF and GUS. Copy number for EHF was divided by copy number of GUS to obtain normalized values for EHF mRNA levels. These normalized values were used for the statistics.
2.3. Statistical analysis Statistical analysis was performed using the SPSS-PC package (Version 17.0; Chicago, IL). Probability of less than .05 was considered significant. The Mann-Whitney U test was used to analyze the differences in gene expression between OC and MM, as well as to compare expression in solid primary OC and OC effusions. The Mann-Whitney U test was also applied to analyses investigating the association between gene expression and clinicopathologic parameters (effusion site, age, histologic grade, FIGO stage, residual disease volume, previous chemotherapy, response to chemotherapy at diagnosis) in OC. For these analyses, clinicopathologic parameters were grouped as follows: age, ≤60 versus N60 years; histologic grade, 1-2 versus 3; FIGO stage, III versus IV; residual disease volume, ≤1 versus N1 cm; previous chemotherapy, yes versus no; response to chemotherapy at diagnosis, complete versus partial response/ stable disease/progression/allergic or adverse reaction. Survival data were available for all OC patients. Univariate survival analyses of progression-free survival (PFS) and overall survival (OS) were executed using the Kaplan-Meier method and groups compared with the logrank test. For this analysis, expression levels were grouped as low or high based on median values. For patients with more than 1 effusion, expression in the first specimen was analyzed. Multivariate analyses of OS and PFS were performed using the Cox proportional hazard model.
3. Results 3.1. EHF is overexpressed in OC compared to MM EHF mRNA expression in OC and MM effusions was compared using qPCR. Pronounced differences were observed between OC (n = 98) and MM (n = 18; 2 failed tests) effusions, with respect to EHF copy number, with OC effusion expression range of 3 to 4881 (median, 449) versus range of 0 to 5 (median, 1) for MM. Primary OC had still higher expression, with copy number range of 74-21 949, median 1535 (Fig. 2). Three surgical specimens from peritoneal MM patients had similar expression to MM effusions (copy nos. 1-5). Statistical analysis showed significantly higher level of EHF expression in OC effusions compared to MM effusions (P b .001), as well as significantly higher EHF levels in primary OC vs. OC effusions (P b .001). Values of the reference gene GUS showed little variation
500
K. Brenne et al.
Fig. 2 EHF and GUS mRNA expression in OC and MM. qPCR amplification plots for EHF and GUS mRNA in OC and MM effusions, as well as in primary OC. Cases were run in triplicate. EHF mRNA expression is much higher in OC compared to MM specimens (this page, whereas the reference gene GUS is uniformly expressed (next page).
EHF in serous ovarian carcinoma
501
Fig. 2 (continued).
502 across different samples, irrespective of tumor type or anatomic site (Fig. 2). EHF expression in the OC effusions, MM effusions, and primary OC is graphically shown in Fig. 3.
3.2. Higher EHF gene expression in primary diagnosis OC effusions is associated with poor PFS
K. Brenne et al. Table 2 Survival determinants in the OC effusion cohort (n = 93 patients) Parameter
Univariate
Multivariate
Relative risk
Overall survival EHF mRNA Age Grade FIGO stage Residual disease Chemoresponse at diagnosis
P P P P P P
= = = = = b
.501 .121 .686 .346 .037 .001
P P P P P P
= = = = = b
.327 .320 .535 .569 .157 .001
0.962 0.989 0.384 0.325 2.005 17.388
PFS EHF mRNA Age Grade FIGO stage Residual disease Chemoresponse at diagnosis
P P P P P P
= .066 = .639 = .481 = .332 = .021 b .001
P P P P P P
= .033 = .436 = .457 = .611 = .052 b .001
4.528 0.606 0.553 0.259 3.784 17.961
In view of the marked differences among OC effusions with respect to EHF mRNA levels, we investigated whether expression of this mRNA is related to clinicopathologic parameters, including survival. EHF mRNA levels were unrelated to effusion site, age, histologic grade, FIGO stage, residual disease volume, previous chemotherapy, or response to chemotherapy at diagnosis or at first disease recurrence (P N .05; data not shown). The follow-up period for the 93 OC patients with effusions analyzed using qPCR ranged from 1 to 120 (median, 25 months). PFS ranged from 0 to 66 months (median, 4 months), with 38 patients never achieving a disease-free period. At the last follow-up, 1 patient was alive with no evidence of disease, 4 patients were alive without disease, and 87 patients were dead of disease. One patient died of unrelated cause. In univariate survival analysis of the entire OC effusion cohort, higher EHF mRNA expression (grouped as low or high based on median values) was associated with a trend for shorter PFS (P = .066), with no relationship to OS. Residual disease volume and chemoresponse at diagnosis were the clinicopathologic parameters that were significantly associated with both OS and PFS in univariate analysis (Table 2). The parameters entered in the Cox multivariate analysis of OS and PFS were EHF mRNA level, age, histologic grade,
FIGO stage, residual disease volume, and response to chemotherapy at diagnosis. In Cox multivariate analysis of OS, chemoresponse at diagnosis was the only independent prognostic factor for OS (P b .001), whereas chemoresponse at diagnosis and EHF mRNA expression were independent predictors of PFS (P b .001 and P = .033, respectively; Table 2). In separate survival analysis of 45 patients with primary diagnosis pre-chemotherapy OC effusions, higher EHF mRNA expression was significantly associated with shorter PFS (P = .01; Fig. 4), with no relationship to OS. As in the analysis of the entire cohort, residual disease volume and
Fig. 3 EHF mRNA expression in OC effusions, primary OC, and MM effusions. EHF mRNA copy number normalized against the housekeeping gene GUS for the different tumor types; primary OC (n = 28), OC/PPC effusions (n = 98), and MM effusions (n = 14). One primary OC sample had a copy number of 21,949 and is not visible on the boxplot that is scaled up to 6000.
Fig. 4 High EHF mRNA expression in OC effusions is associated with poor survival. Kaplan-Meier survival curve showing the association between EHF mRNA expression and PFS for women with pre-chemotherapy effusions (n = 45). Patients with effusions with high (above median) EHF mRNA expression levels (n = 22; dashed line) had mean PFS of 3 months compared with 12 months for patients with effusions having low EHF mRNA expression levels (n = 23, solid line; P = .01).
Bold indicates statistically significant values (P b.05).
EHF in serous ovarian carcinoma
503
chemoresponse at diagnosis were significantly associated with both OS and PFS in univariate analysis (Table 3). In Cox multivariate analysis of OS for patients with prechemotherapy effusions, chemoresponse at diagnosis (P = .001) and residual disease volume (P = .026) were independent prognostic factors, whereas chemoresponse at diagnosis (P b .001), residual disease volume (P = .016), and EHF mRNA expression (P = .009) were independent predictors of PFS (Table 3). EHF mRNA expression was unrelated to survival in separate analysis of post-chemotherapy effusions.
4. Discussion Despite recent improvements in the IHC panels used to identify the origin and histologic type of tumors, there are still difficulties encountered in effusion cytology and surgical pathology. In the anatomical context of the serosal cavities, these difficulties most frequently involve differentiating serous OC/PPC from MM, and determining the site of origin for metastatic adenocarcinoma [11,14-18]. As OC/ PPC and MM present with similar clinical symptoms (especially within the peritoneal cavity), morphological resemblance, and co-expression or lack of expression of several molecular markers, these 2 tumor types are especially difficult to discriminate from one another. Examination of surgical specimens has shown that podoplanin/D2-40, calretinin, cytokeratin 5/6, WT1, and mesothelin are markers that are expressed in both OC/PPC and MM, whereas both tumors are carcinoembryonic antigen-negative (reviewed in references [15,17]). These findings have later been shown to apply to effusion cytology as well [19,20]. Table 3 Survival determinants for patients with primary diagnosis pre-chemotherapy OC effusions (n = 45) Parameter
Univariate
Multivariate
Relative risk
Overall survival EHF mRNA Age Grade FIGO stage Residual disease Chemoresponse at diagnosis
P P P P P P
= .332 = .301 = .304 = .589 = .006 b .001
P P P P P P
= .195 = .964 = .238 = .795 = .026 = .001
1.678 0.002 1.395 0.068 4.989 10.371
PFS EHF mRNA Age Grade FIGO stage Residual disease Chemoresponse at diagnosis
P P P P P P
= .01 = .901 = .261 = .475 = .003 b .001
P P P P P P
= .009 = .322 = .270 = .704 = .016 b .001
6.879 0.983 1.218 0.145 5.809 22.332
Bold indicates statistically significant values (P b.05).
Ets transcription factors are expressed in multiple cancers, and their role in differentiating tumors of various origin has consequently not been a central research topic to date. Elevated expression levels of EHF mRNA have previously been detected in breast and colorectal carcinoma [21-23], whereas decreased expression of EHF mRNA has been observed in prostate and head and neck carcinoma [24,25]. With the exception of the previous cDNA microarray study by our group [12], EHF expression has not been studied in OC to date. In the present study, we measured the gene expression levels of EHF in OC and MM effusions, and observed significantly higher expression levels of EHF mRNA in the former tumor. These results concur with those of our previously published cDNA microarray analysis, in which EHF was found to be overexpressed in this tumor compared to peritoneal MM [12]. Low to negative expression of EHF was detected in the MM effusions, whereas OC effusions all had distinct expression of the gene. qPCR is increasingly used in routine diagnostics. The method is fast and accurate, and allows for relatively consistent and objective interpretation of results. Challenges associated with qPCR are a demanding validation process, the need for skilled users and expensive equipment. The use of RNA as sample material requires special handling, including freezing of fresh material and sample handling in an RNasefree environment. Despite these challenges, we believe that a diagnostic qPCR panel may effectively resolve cases with equivocal effusion cytology. The clear-cut difference in EHF mRNA expression between OC and MM provides strong evidence in favor of including this assay in such qPCR panel, although this requires validation in independent series. We plan to test the strength of the EHF assay in the near future as part of a multigene assay including the folate receptors genes FOLR1 and FOLR3, which are overexpressed in OC, and the tenascin X gene TNXB, overexpressed in MM, which we recently reported to be powerful differentiators between these tumors using qPCR [26,27], as well as other genes currently evaluated by our group. When comparing the gene expression levels of EHF between primary OC and OC effusions, we found significantly higher EHF expression at the former anatomic site. From these results one can infer that there is decreased expression of EHF as the tumor progresses and metastasizes from its solid form to effusions. The altered expression profile of tumor cells in effusions may result from a change in the microenvironment. When tumor cells are found in solid organs, they are complemented by various host cell populations such as stromal myofibroblasts, endothelial cells, and leukocytes. Within solid organs, tumor cells are able to induce leaky vessels and obtain nutrients and oxygen, in addition to gaining access to the circulation. When present in effusions, however, tumor cells are in a unique microenvironment with reduced direct access to oxygen and nutrients, and no longer in interaction with stromal myofibroblasts and endothelial cells [28]. The decreased
504 expression levels of EHF in OC effusions compared to primary OC may be the result of such change in the microenvironment. As Ets family members are involved in angiogenesis, this reduced expression may also be related to the absent need to form new vessels when tumor cells grow in suspension. Previous studies by our group have documented differences in the expression and clinical significance of cancerassociated molecules between OC cells in pre- and postchemotherapy effusions [28,29]. Thus, analysis of the relationship between EHF mRNA expression in OC/PPC effusions and previous exposure to chemotherapy, as well as to chemoresponse at diagnosis and at first disease recurrence, was conducted. No difference in EHF expression was found between pre- and post-chemotherapy specimens, nor were the levels of this gene related to chemotherapy response. This suggests that the mRNA expression of EHF in OC/PPC effusions is unaffected by chemotherapy. When examining the prognostic role of EHF mRNA expression in OC, we found a trend for shorter PFS for patients with effusions expressing high EHF mRNA levels, which became significant when the analysis was limited to patients with pre-chemotherapy effusions. In agreement with several of our previously published studies [30-33], a significant association was additionally observed between residual disease volume and response to chemotherapy at diagnosis and PFS in both patient groups. In Cox multivariate analysis of the entire cohort, EHF mRNA expression and chemoresponse at diagnosis were independent predictors of shorter PFS, whereas all 3 parameters, that is, EHF mRNA expression, chemoresponse at diagnosis, and residual disease volume, were independent predictors of poor PFS for patients with pre-chemotherapy effusions. This suggests that EHF mRNA expression may be a novel prognostic marker in OC. This finding is in full agreement with our previous studies of Ets family members in OC, in which mRNA expression of Ets-1 and PEA3, 2 additional members of this transcription factor family, was associated with poor survival [34-36]. One should nevertheless note that the follow-up period in the present study was short because most of the patients were diagnosed at stage IIIc or IV and therefore did not survive longer. Analysis of a separate cohort with more balanced FIGO stage distribution and longer follow-up may therefore be required for definite conclusions, and we plan to do so in the near future. Another potential limitation of our cohort is the fact that residual disease data were absent for some of the patients, reflecting the fact the some of these advanced-stage OC patients were inoperable, as well as the fact that some patients were primarily operated on at other hospitals and referred to our hospital with recurrent disease. In conclusion, analysis of EHF mRNA expression in a large cohort of patients with primary OC, OC effusions, and MM effusions showed increased levels of this gene in OC effusions compared to MM effusions, supporting the use of the EHF mRNA assay as a novel diagnostic test in this
K. Brenne et al. setting. In addition, EHF mRNA levels were higher in primary OC compared to OC effusions. The biological rationale for this difference is yet to be established. Finally, higher EHF mRNA expression in OC effusions from patients with advanced disease correlated with poor survival, encouraging further research into the relationship between EHF mRNA levels and disease outcome in this cancer.
References [1] Oikawa T, Yamada T. Molecular biology of the Ets family of transcription factors. Gene 2003;303:11-34. [2] Oikawa T. ETS transcription factors: possible targets for cancer therapy. Cancer Sci 2004;95:626-33. [3] Seth A, Watson DK. ETS transcription factors and their emerging roles in human cancer. Eur J Cancer 2005;41:2462-78. [4] Hsu T, Trojanowska M, Watson DK. Ets proteins in biological control and cancer. J Cell Biochem 2004;91:896-903. [5] Sementchenko VI, Watson DK. Ets target genes: past, present and future. Oncogene 2000;19:6533-48. [6] Dittmer J. The biology of the Ets 1 proto-oncogene. Mol Cancer 2003;2:29. [7] Laudet V, Hanni C, Stehelin D, et al. Molecular phylogeny of the ETS gene family. Oncogene 1999;18:1351-9. [8] Kas K, Finger E, Grall F. ESE-3, a novel member of an epitheliumspecific Ets transcription factor subfamily, demonstrates different target gene specificity from ESE-1. J Biol Chem 2000;275:2986-98. [9] American Cancer Society. Cancer facts and figures 2010. Estimated new cancers cases and deaths by sex for all sites. Available at: http:// www.cancer.org/Research/CancerFactsFigures/CancerFactsFigures/ cancer-facts-and-figures-2010 (Accessed July 31, 2010). [10] Davidson B. Malignant effusions: from diagnosis to biology. Diagn Cytopathol 2004;31:246-54. [11] Davidson B. Biological characteristics of cancers involving the serosal cavities. Crit Rev Oncog 2007;13:189-227. [12] Davidson B, Zhang Z, Kleinberg L, et al. Gene expression signatures differentiate ovarian/peritoneal serous carcinoma from diffuse malignant peritoneal mesothelioma. Clin Cancer Res 2006;12:5944-50. [13] Beillard E, Pallisgaard N, van der Velden VH, et al. Evaluation of candidate control genes for diagnosis and residual disease detection in leukemic patients using “real-time” quantitative reverse-transcriptase polymerase reaction (RQ-PCR)—a Europe against cancer program. Leukemia 2003;17:2474-86. [14] Marchevsky AM. Application of immunohistochemistry to the diagnosis of malignant mesothelioma. Arch Pathol Lad Med 2008; 132:397-401. [15] Ordóñez NG. Value of immunohistochemistry in distinguishing peritoneal mesothelioma from serous carcinoma of the ovary and peritoneum: a review and update. Adv Anat Pathol 2006;13:16-25. [16] Ordóñez NG. What are the current best immunohistochemical markers for the diagnosis of epithelioid mesothelioma? A review and update. HUM PATHOL 2007;38:1-16. [17] Suster S, Moran CA. Applications and limitations of immunohistochemistry in the diagnosis of malignant mesothelioma. Adv Anat Pathol 2006;13:316-29. [18] Fetsch PA, Abati A. Immunocytochemistry in effusion cytology. Cancer 2001;93:293-308. [19] Davidson B, Nielsen S, Christensen J, et al. The role of desmin and Ncadherin in effusion cytology. Am J Surg Pathol 2001;25:1405-12. [20] Bassarova AV, Nesland JM, Davidson B. D2-40 is not a specific marker for cells of mesothelial origin in serous effusions. Am J Surg Pathol 2006;30:878-82. [21] He J, Pan Y, Hu J, et al. Profile of ETS gene expression in human breast carcinoma. Cancer Biol Ther 2007;6:76-82.
EHF in serous ovarian carcinoma [22] Galang CK, Mullers WJ, Foos G, et al. Changes in the expression of many ETS family transcription factors and of potential target genes in normal mammary tissue and tumors. J Biol Chem 2004;19:11281-92. [23] Hufton SE, Moerkerk PT, Brandtwijk R, et al. A profile of differentially expressed genes in primary colorectal cancer using suppression subtractive hybridization. FEBS Lett 1999;463:77-82. [24] Cangemi R, Mensah A, Albertini V, et al. Reduced expression and tumor suppressor function of the ETS transcription factor ESE-3 in prostate cancer. Oncogene 2008;27:2877-85. [25] Gonzalez HE, Gujrati M, Frederick M, et al. Identification of 9 genes differentially expressed in head and neck squamous cell carcinoma. Arch Otolaryngol Head Neck Surg 2003;129:754-9. [26] Yuan Y, Nymoen DA, Dong HP, et al. Expression of the folate receptor genes FOLR1 and FOLR3 differentiates ovarian carcinoma from breast carcinoma and malignant mesothelioma in serous effusions. HUM PATHOL 2009;40:1453-60. [27] Yuan Y, Nymoen DA, Stavnes HT, et al. Tenascin-X is a novel diagnostic marker of malignant mesothelioma. Am J Surg Pathol 2009;33:1673-82. [28] Davidson B. Anatomic site-related expression of cancer-associated molecules in ovarian carcinoma. Curr Cancer Drug Targets 2007;7: 109-20.
505 [29] Davidson B, Espina V, Steinberg AM, et al. Proteomic analysis of malignant ovarian cancer effusions as a tool for biologic and prognostic profiling. Clin Cancer Res 2006;12:791-9. [30] Cannistra SA. Cancer of the ovary. N Engl J Med 2004;351:2519-29. [31] Aletti GD, Gallenberg MD, Cliby WA, et al. Current management strategies for ovarian cancer. Mayo Clin Proc 2007;82:751-70. [32] Hoskins WI, McGuire WP, Brady MF, et al. The effect of diameter of largest residual disease on survival after primary cytoreductive surgery in patients with suboptimal residual epithelial ovarian carcinoma. Am J Obstet Gynecol 1994;170:974-9. [33] Rose PG, Nerenstone S, Brady MF, et al. Secondary surgical cytoreduction for advanced ovarian carcinoma. N Engl J Med 2004; 351:2489-546. [34] Davidson B, Reich R, Goldberg I, et al. Ets-1 messenger RNA expression is a novel marker of poor survival in ovarian carcinoma. Clin Cancer Res 2001;7:551-7. [35] Davidson B, Goldberg I, Gotlieb WH, et al. PEA3 is the second Ets family transcription factor involved in tumor progression in ovarian carcinoma. Clin Cancer Res 2003;9:1412-9. [36] Davidson B, Risberg B, Goldberg I, et al. Ets-1 mRNA expression in effusions of serous ovarian carcinoma patients is a marker of poor outcome. Am J Surg Pathol 2001;25:1493-500.
www.elsevier.com/locate/humpath
Original contribution
Expression of the Ets transcription factor EHF in serous ovarian carcinoma effusions is a marker of poor survival☆ Kjersti Brenne MSc a , Dag André Nymoen MSc a , Thea Eline Hetland MD b , Claes G. Trope' MD, PhD b,c , Ben Davidson MD, PhD a,c,⁎ a
Division of Pathology, Norwegian Radium Hospital, Oslo University Hospital, N-0424 Oslo, Norway Department of Gynecologic Oncology, Norwegian Radium Hospital, Oslo University Hospital, N-0424 Oslo, Norway c University of Oslo, Faculty of Medicine, Institute of Clinical Medicine, N-0424 Oslo, Norway b
Received 22 January 2011; revised 19 May 2011; accepted 20 May 2011
Keywords: Ovarian carcinoma; Malignant mesothelioma; Diagnosis; Survival; Quantitative real-time PCR
Summary The EHF (Ets homologous factor) gene was previously shown to be overexpressed in ovarian/ primary peritoneal serous carcinoma compared to malignant mesothelioma using gene expression arrays. The objective of this study was to validate this finding at the mRNA level in a larger series. We analyzed the diagnostic role of EHF in 98 ovarian serous carcinoma effusions, 23 malignant mesothelioma specimens (20 effusions, 3 surgical specimens), and 28 primary ovarian serous carcinomas using quantitative real-time polymerase chain reaction. Expression levels of EHF in ovarian carcinoma were additionally investigated for association with clinicopathologic parameters and survival. Quantitative real-time polymerase chain reaction analysis showed significantly higher expression of EHF mRNA in ovarian carcinoma effusions and in primary ovarian carcinoma compared to malignant mesothelioma effusions (P b .001 for both). EHF mRNA expression was additionally higher in primary ovarian carcinomas compared to effusions of this cancer (P b .001). In univariate analysis for all patients with effusions, higher EHF mRNA levels were associated with a trend for shorter progression-free survival (P = .066), which became significant in analysis of 45 patients with primary diagnosis pre-chemotherapy effusions (P = .01). In Cox multivariate analysis, EHF mRNA expression was an independent predictor of poor progression-free survival for all patients and patients with primary diagnosis pre-chemotherapy effusions (P = .033 and P = .009, respectively). EHF mRNA levels differentiate ovarian carcinoma from malignant mesothelioma and may thus be of diagnostic value in this setting. EHF may be a novel prognostic marker in ovarian carcinoma. © 2012 Elsevier Inc. All rights reserved.
Abbreviations: EHF, Ets homology factor; qPCR, quantitative real-time PCR; EBS, Ets binding site; ESE-3, epithelial specific ETS-3; OC, ovarian carcinoma; PPC, primary peritoneal carcinoma; MM, malignant mesothelioma; FIGO, International Federation of Gynecology and Obstetrics; H&E, hematoxylin and eosin; PFS, progression-free survival; OS, overall survival. ☆ This work was supported by grants from the Norwegian Cancer Society and the Research Fund at the Norwegian Radium Hospital, and by the Inger and John Fredriksen Foundation for Ovarian Cancer Research. ⁎ Corresponding author. Department of Pathology, Norwegian Radium Hospital, Oslo University Hospital, Montebello N-0310 Oslo, Norway. E-mail address: [email protected] (B. Davidson). 0046-8177/$ – see front matter © 2012 Elsevier Inc. All rights reserved. doi:10.1016/j.humpath.2011.05.023
EHF in serous ovarian carcinoma
1. Introduction Ets homologous factor (EHF) is member of the ETS gene family, one of the largest families of transcriptional regulators. The ETS gene family was discovered through the identification of the viral oncogene v-ets in the avian transforming retrovirus E26, and is divided into subfamilies (ELF, ELG, ERG, ERF, ESE, ETS, PEA3, ER71, SPI, TCF, TEL) based on sequence similarity in the ETS domain. The ETS domain is an evolutionarily conserved winged helix-turn-helix DNA binding domain consisting of 85 amino acids that is present in all ETS proteins. This DNA binding domain recognizes a core GGAA/T sequence known as the Ets binding site (EBS) [1-7]. All ETS transcription factors bind to the EBS, and such EBS have been identified in the promoter and enhancer regions of viral and cellular genes [5]. ETS gene products activate or repress expression of genes that are involved in cellular proliferation, differentiation, hematopoiesis, apoptosis, tissue remodeling, angiogenesis, transformation, invasion and metastasis. EHF, also known as ESE-3 (epithelial specific ETS-3), is a member of the ESE subfamily which is epithelial-specific and has distinct functions in epithelial cell differentiation and proliferation [8]. As a result of its involvement in these cellular processes, uncoordinated expression of EHF has the potential to contribute to tumor growth and consequently disease progression. Ovarian carcinoma (OC) is the second most common gynecologic cancer, and has the highest death toll among gynecologic cancers, with more than 50% of diagnosed women dying of their disease [9]. The poor outcome of OC reflects the fact that most patients present with advanced stage (III/IV) disease. Despite efforts to improve disease outcome by focus on early detection and the combination of aggressive surgery with improved chemotherapy protocols, the mortality rate of OC has remained largely unchanged. The serosal cavities, including the peritoneal, pleural, and pericardial spaces, are frequently affected by cancer. The presence of cancer cells in effusions at these anatomic sites is evidence of advanced-stage disease with metastatic spread, and regardless of the tumor site of origin, marks disease progression and is associated with poor survival [10,11]. One of the most challenging differential diagnoses in serosal cancer is between ovarian/primary peritoneal serous carcinoma (OC/PPC) and malignant mesothelioma (MM), tumors that are closely related in terms of clinical presentation, morphology, and immunohistochemical phenotype. We recently reported on the differential expression of 189 genes in OC/PPC compared with diffuse peritoneal MM using cDNA microarray technology [12]. Among the differentially expressed genes, EHF was identified as a gene that is significantly overexpressed in OC/PPC effusions compared to peritoneal MM effusions. The present study evaluated the diagnostic value of EHF mRNA expression
497 levels in differentiating OC from MM and analyzed the prognostic value of EHF in OC.
2. Materials and methods 2.1. Patients and material 2.1.1. Effusions The material analyzed in the present study consisted of 112 fresh-frozen effusions, including 98 Müllerian carcinomas and 23 MMs, submitted for routine diagnostic purposes to the Division of Pathology, Norwegian Radium Hospital, during the period 1998 to 2005. Müllerian carcinoma effusions (74 peritoneal, 24 pleural) were obtained from 93 patients (5 patients with 2 effusions each) diagnosed with International Federation of Gynecology and Obstetrics (FIGO) stage II-IV OC (n = 77), PPC (n = 11), or carcinoma of the fallopian tube (n = 5). Most of the specimens (82/98; 84%) were of the serous type (Fig. 1A-C). Because of their closely linked histogenesis and phenotype, all these tumors are referred to as OC in the following sections. Clinicopathologic data for patients with OC effusions were obtained from the Department of Gynecologic Oncology at the Norwegian Radium Hospital and are summarized in Table 1. Most of the patients (83/93; 89%) received platinum-based chemotherapy at diagnosis. The 23 MM specimens consisted of 20 effusions (15 pleural, 5 peritoneal) (Fig. 1D-F) and 3 surgical specimens of diffuse peritoneal MM. All were from patients diagnosed with MM of the epithelioid or biphasic type in biopsy specimens. Effusion specimens were centrifuged immediately after tapping, and cell pellets were frozen at −70°C in equal amounts of RPMI 1640 medium containing 50% fetal calf serum and 20% dimethylsulfoxide. Smears and hematoxylin and eosin (H&E)–stained cell block sections were reviewed by a surgical pathologist experienced in cytopathology (B. D.). Diagnoses were based on morphology and immunohistochemistry. The antibody panel used for differentiating OC/PPC from MM and for establishing the malignant character of the mesothelial cell proliferations included Ber-EP4, B72.3, calretinin, desmin, and EMA. BG8 was added as a carcinoma marker in some of the cases, whereas WT-1 was used as marker of serous differentiation in OC specimens. The tumor cell population was more than 50% in all specimens analyzed in this study. 2.1.2. Primary OC In addition to the above-mentioned effusions, 28 solid primary OC, all of the serous type, were examined (Fig. 1G-I). Histologic grade was as follows: grade 1, five tumors; grade 2, five tumors; grade 3, eighteen tumors. FIGO stage was as follows: I, 1 patient; II, 1 patient; III, 20 patients; IV, 6 patients. Residual disease volume was available for 24 patients, of whom 16 were debulked to ≤1
498
K. Brenne et al.
Fig. 1 Morphology. Examples of 3 serous carcinoma effusions (A-C), 3 MM effusions (D-F), and 3 primary ovarian serous carcinomas (G-I); A and D, PAP stain; B and E, Diff-Quik stain; C, F to I, H&E.
cm and 8 to N1 cm. Surgical specimens were snap-frozen and kept at −70°C with no medium. Frozen sections from all tumors were evaluated for the presence of a N50% tumor component and absence of necrosis. H&E-stained sections from these tumors were reviewed to establish tumor type and histologic grade. The same procedure was applied to the 3 solid MM specimens.
2.2. Quantitative real-time PCR Total RNA was extracted using the automated sample preparation instrument QiaCube and the RNeasy Mini Kit
(Qiagen, Hilden, Germany). mRNA was reverse-transcribed into cDNA using Superscript III Reverse Transcriptase (Invitrogen, Carlsbad, CA). The EHF (NM_012153.3) assay covered exon junction 12. Primer specificity and efficiency were validated by gel electrophoresis and Power SYBR Green (Applied Biosystems, Foster City, CA), respectively. The assay was controlled for primer dimers using the NetPrimer software by PREMIER Biosoft, as well as for single nucleotide polymorphisms through the NCBI database. Primer efficiency was tested using Power SYBR Green (Applied Biosystems) and a dilution series of synthetic oligonucleotide as template and subsequently as standard curve. The quantitative real-time
EHF in serous ovarian carcinoma
499
Table 1 Clinicopathologic data of the OC effusion cohort (n = 93 patients) Parameter
No. of patients
Age
34-87 y (mean = 61) 1 53 39 7 20 56 10 78 11 4 32 47 14 49 40 4
FIGO stage
Grade
Histology
Residual disease
Chemoresponse at diagnosis
II III IV I II III NA a Serous Non-serous b NA c ≤1 cm N1 cm NA d Complete Incomplete e ND f
a NA = nonavailable; including effusions from inoperable patients where biopsy was too small for grading and patients operated in other hospitals, for which the primary tumor could not be accessed for assessment of grade. b Including clear cell, endometrioid, undifferentiated, and mixed histology carcinomas. c Patients with histologically diagnosed OC for which the primary tumor could not be accessed for assessment of tumor type. d Patients who were inoperable or had no record. e Partial response/stable disease/progression/allergic or adverse reaction. f ND = nondetermined, including patients who received no chemotherapy and patients who died before chemoresponse could be assessed.
PCR (qPCR) reaction was run using the Platinum qPCR SuperMix-UDG with ROX solution (Invitrogen) and quantified on the Applied Biosystems 7900HT Sequence Detection System. Primer and probe sequences as follows: Forward primer 5′- ACCCAGAATCTTTAGGTAAATGAGATC- 3′ Reverse primer 5′- CTGGTGAAGGAGGTTGTTGC-3′ Probe Fam 5′- ATTCTGGAAGGAGGTGGTGTAATGAATCTCAACC-3′ nonfluorescent quencher
An array of 12 reference genes (TaqMan low-density array human endogenous control panel; Applied Biosystems) was tested to identify the most uniformly expressed transcript in effusions specimens. Based on these results, the expression levels of EHF were normalized against the housekeeping gene β-glucuronidase (GUS). The GUS primer and probe sequences have been published elsewhere [13]. Standard curves for the GUS assay were purchased from Ipsogen (Marseille, France).
Synthetic oligonucleotides were used as standard curve for EHF. Standard curves were used to obtain copy number of EHF and GUS. Copy number for EHF was divided by copy number of GUS to obtain normalized values for EHF mRNA levels. These normalized values were used for the statistics.
2.3. Statistical analysis Statistical analysis was performed using the SPSS-PC package (Version 17.0; Chicago, IL). Probability of less than .05 was considered significant. The Mann-Whitney U test was used to analyze the differences in gene expression between OC and MM, as well as to compare expression in solid primary OC and OC effusions. The Mann-Whitney U test was also applied to analyses investigating the association between gene expression and clinicopathologic parameters (effusion site, age, histologic grade, FIGO stage, residual disease volume, previous chemotherapy, response to chemotherapy at diagnosis) in OC. For these analyses, clinicopathologic parameters were grouped as follows: age, ≤60 versus N60 years; histologic grade, 1-2 versus 3; FIGO stage, III versus IV; residual disease volume, ≤1 versus N1 cm; previous chemotherapy, yes versus no; response to chemotherapy at diagnosis, complete versus partial response/ stable disease/progression/allergic or adverse reaction. Survival data were available for all OC patients. Univariate survival analyses of progression-free survival (PFS) and overall survival (OS) were executed using the Kaplan-Meier method and groups compared with the logrank test. For this analysis, expression levels were grouped as low or high based on median values. For patients with more than 1 effusion, expression in the first specimen was analyzed. Multivariate analyses of OS and PFS were performed using the Cox proportional hazard model.
3. Results 3.1. EHF is overexpressed in OC compared to MM EHF mRNA expression in OC and MM effusions was compared using qPCR. Pronounced differences were observed between OC (n = 98) and MM (n = 18; 2 failed tests) effusions, with respect to EHF copy number, with OC effusion expression range of 3 to 4881 (median, 449) versus range of 0 to 5 (median, 1) for MM. Primary OC had still higher expression, with copy number range of 74-21 949, median 1535 (Fig. 2). Three surgical specimens from peritoneal MM patients had similar expression to MM effusions (copy nos. 1-5). Statistical analysis showed significantly higher level of EHF expression in OC effusions compared to MM effusions (P b .001), as well as significantly higher EHF levels in primary OC vs. OC effusions (P b .001). Values of the reference gene GUS showed little variation
500
K. Brenne et al.
Fig. 2 EHF and GUS mRNA expression in OC and MM. qPCR amplification plots for EHF and GUS mRNA in OC and MM effusions, as well as in primary OC. Cases were run in triplicate. EHF mRNA expression is much higher in OC compared to MM specimens (this page, whereas the reference gene GUS is uniformly expressed (next page).
EHF in serous ovarian carcinoma
501
Fig. 2 (continued).
502 across different samples, irrespective of tumor type or anatomic site (Fig. 2). EHF expression in the OC effusions, MM effusions, and primary OC is graphically shown in Fig. 3.
3.2. Higher EHF gene expression in primary diagnosis OC effusions is associated with poor PFS
K. Brenne et al. Table 2 Survival determinants in the OC effusion cohort (n = 93 patients) Parameter
Univariate
Multivariate
Relative risk
Overall survival EHF mRNA Age Grade FIGO stage Residual disease Chemoresponse at diagnosis
P P P P P P
= = = = = b
.501 .121 .686 .346 .037 .001
P P P P P P
= = = = = b
.327 .320 .535 .569 .157 .001
0.962 0.989 0.384 0.325 2.005 17.388
PFS EHF mRNA Age Grade FIGO stage Residual disease Chemoresponse at diagnosis
P P P P P P
= .066 = .639 = .481 = .332 = .021 b .001
P P P P P P
= .033 = .436 = .457 = .611 = .052 b .001
4.528 0.606 0.553 0.259 3.784 17.961
In view of the marked differences among OC effusions with respect to EHF mRNA levels, we investigated whether expression of this mRNA is related to clinicopathologic parameters, including survival. EHF mRNA levels were unrelated to effusion site, age, histologic grade, FIGO stage, residual disease volume, previous chemotherapy, or response to chemotherapy at diagnosis or at first disease recurrence (P N .05; data not shown). The follow-up period for the 93 OC patients with effusions analyzed using qPCR ranged from 1 to 120 (median, 25 months). PFS ranged from 0 to 66 months (median, 4 months), with 38 patients never achieving a disease-free period. At the last follow-up, 1 patient was alive with no evidence of disease, 4 patients were alive without disease, and 87 patients were dead of disease. One patient died of unrelated cause. In univariate survival analysis of the entire OC effusion cohort, higher EHF mRNA expression (grouped as low or high based on median values) was associated with a trend for shorter PFS (P = .066), with no relationship to OS. Residual disease volume and chemoresponse at diagnosis were the clinicopathologic parameters that were significantly associated with both OS and PFS in univariate analysis (Table 2). The parameters entered in the Cox multivariate analysis of OS and PFS were EHF mRNA level, age, histologic grade,
FIGO stage, residual disease volume, and response to chemotherapy at diagnosis. In Cox multivariate analysis of OS, chemoresponse at diagnosis was the only independent prognostic factor for OS (P b .001), whereas chemoresponse at diagnosis and EHF mRNA expression were independent predictors of PFS (P b .001 and P = .033, respectively; Table 2). In separate survival analysis of 45 patients with primary diagnosis pre-chemotherapy OC effusions, higher EHF mRNA expression was significantly associated with shorter PFS (P = .01; Fig. 4), with no relationship to OS. As in the analysis of the entire cohort, residual disease volume and
Fig. 3 EHF mRNA expression in OC effusions, primary OC, and MM effusions. EHF mRNA copy number normalized against the housekeeping gene GUS for the different tumor types; primary OC (n = 28), OC/PPC effusions (n = 98), and MM effusions (n = 14). One primary OC sample had a copy number of 21,949 and is not visible on the boxplot that is scaled up to 6000.
Fig. 4 High EHF mRNA expression in OC effusions is associated with poor survival. Kaplan-Meier survival curve showing the association between EHF mRNA expression and PFS for women with pre-chemotherapy effusions (n = 45). Patients with effusions with high (above median) EHF mRNA expression levels (n = 22; dashed line) had mean PFS of 3 months compared with 12 months for patients with effusions having low EHF mRNA expression levels (n = 23, solid line; P = .01).
Bold indicates statistically significant values (P b.05).
EHF in serous ovarian carcinoma
503
chemoresponse at diagnosis were significantly associated with both OS and PFS in univariate analysis (Table 3). In Cox multivariate analysis of OS for patients with prechemotherapy effusions, chemoresponse at diagnosis (P = .001) and residual disease volume (P = .026) were independent prognostic factors, whereas chemoresponse at diagnosis (P b .001), residual disease volume (P = .016), and EHF mRNA expression (P = .009) were independent predictors of PFS (Table 3). EHF mRNA expression was unrelated to survival in separate analysis of post-chemotherapy effusions.
4. Discussion Despite recent improvements in the IHC panels used to identify the origin and histologic type of tumors, there are still difficulties encountered in effusion cytology and surgical pathology. In the anatomical context of the serosal cavities, these difficulties most frequently involve differentiating serous OC/PPC from MM, and determining the site of origin for metastatic adenocarcinoma [11,14-18]. As OC/ PPC and MM present with similar clinical symptoms (especially within the peritoneal cavity), morphological resemblance, and co-expression or lack of expression of several molecular markers, these 2 tumor types are especially difficult to discriminate from one another. Examination of surgical specimens has shown that podoplanin/D2-40, calretinin, cytokeratin 5/6, WT1, and mesothelin are markers that are expressed in both OC/PPC and MM, whereas both tumors are carcinoembryonic antigen-negative (reviewed in references [15,17]). These findings have later been shown to apply to effusion cytology as well [19,20]. Table 3 Survival determinants for patients with primary diagnosis pre-chemotherapy OC effusions (n = 45) Parameter
Univariate
Multivariate
Relative risk
Overall survival EHF mRNA Age Grade FIGO stage Residual disease Chemoresponse at diagnosis
P P P P P P
= .332 = .301 = .304 = .589 = .006 b .001
P P P P P P
= .195 = .964 = .238 = .795 = .026 = .001
1.678 0.002 1.395 0.068 4.989 10.371
PFS EHF mRNA Age Grade FIGO stage Residual disease Chemoresponse at diagnosis
P P P P P P
= .01 = .901 = .261 = .475 = .003 b .001
P P P P P P
= .009 = .322 = .270 = .704 = .016 b .001
6.879 0.983 1.218 0.145 5.809 22.332
Bold indicates statistically significant values (P b.05).
Ets transcription factors are expressed in multiple cancers, and their role in differentiating tumors of various origin has consequently not been a central research topic to date. Elevated expression levels of EHF mRNA have previously been detected in breast and colorectal carcinoma [21-23], whereas decreased expression of EHF mRNA has been observed in prostate and head and neck carcinoma [24,25]. With the exception of the previous cDNA microarray study by our group [12], EHF expression has not been studied in OC to date. In the present study, we measured the gene expression levels of EHF in OC and MM effusions, and observed significantly higher expression levels of EHF mRNA in the former tumor. These results concur with those of our previously published cDNA microarray analysis, in which EHF was found to be overexpressed in this tumor compared to peritoneal MM [12]. Low to negative expression of EHF was detected in the MM effusions, whereas OC effusions all had distinct expression of the gene. qPCR is increasingly used in routine diagnostics. The method is fast and accurate, and allows for relatively consistent and objective interpretation of results. Challenges associated with qPCR are a demanding validation process, the need for skilled users and expensive equipment. The use of RNA as sample material requires special handling, including freezing of fresh material and sample handling in an RNasefree environment. Despite these challenges, we believe that a diagnostic qPCR panel may effectively resolve cases with equivocal effusion cytology. The clear-cut difference in EHF mRNA expression between OC and MM provides strong evidence in favor of including this assay in such qPCR panel, although this requires validation in independent series. We plan to test the strength of the EHF assay in the near future as part of a multigene assay including the folate receptors genes FOLR1 and FOLR3, which are overexpressed in OC, and the tenascin X gene TNXB, overexpressed in MM, which we recently reported to be powerful differentiators between these tumors using qPCR [26,27], as well as other genes currently evaluated by our group. When comparing the gene expression levels of EHF between primary OC and OC effusions, we found significantly higher EHF expression at the former anatomic site. From these results one can infer that there is decreased expression of EHF as the tumor progresses and metastasizes from its solid form to effusions. The altered expression profile of tumor cells in effusions may result from a change in the microenvironment. When tumor cells are found in solid organs, they are complemented by various host cell populations such as stromal myofibroblasts, endothelial cells, and leukocytes. Within solid organs, tumor cells are able to induce leaky vessels and obtain nutrients and oxygen, in addition to gaining access to the circulation. When present in effusions, however, tumor cells are in a unique microenvironment with reduced direct access to oxygen and nutrients, and no longer in interaction with stromal myofibroblasts and endothelial cells [28]. The decreased
504 expression levels of EHF in OC effusions compared to primary OC may be the result of such change in the microenvironment. As Ets family members are involved in angiogenesis, this reduced expression may also be related to the absent need to form new vessels when tumor cells grow in suspension. Previous studies by our group have documented differences in the expression and clinical significance of cancerassociated molecules between OC cells in pre- and postchemotherapy effusions [28,29]. Thus, analysis of the relationship between EHF mRNA expression in OC/PPC effusions and previous exposure to chemotherapy, as well as to chemoresponse at diagnosis and at first disease recurrence, was conducted. No difference in EHF expression was found between pre- and post-chemotherapy specimens, nor were the levels of this gene related to chemotherapy response. This suggests that the mRNA expression of EHF in OC/PPC effusions is unaffected by chemotherapy. When examining the prognostic role of EHF mRNA expression in OC, we found a trend for shorter PFS for patients with effusions expressing high EHF mRNA levels, which became significant when the analysis was limited to patients with pre-chemotherapy effusions. In agreement with several of our previously published studies [30-33], a significant association was additionally observed between residual disease volume and response to chemotherapy at diagnosis and PFS in both patient groups. In Cox multivariate analysis of the entire cohort, EHF mRNA expression and chemoresponse at diagnosis were independent predictors of shorter PFS, whereas all 3 parameters, that is, EHF mRNA expression, chemoresponse at diagnosis, and residual disease volume, were independent predictors of poor PFS for patients with pre-chemotherapy effusions. This suggests that EHF mRNA expression may be a novel prognostic marker in OC. This finding is in full agreement with our previous studies of Ets family members in OC, in which mRNA expression of Ets-1 and PEA3, 2 additional members of this transcription factor family, was associated with poor survival [34-36]. One should nevertheless note that the follow-up period in the present study was short because most of the patients were diagnosed at stage IIIc or IV and therefore did not survive longer. Analysis of a separate cohort with more balanced FIGO stage distribution and longer follow-up may therefore be required for definite conclusions, and we plan to do so in the near future. Another potential limitation of our cohort is the fact that residual disease data were absent for some of the patients, reflecting the fact the some of these advanced-stage OC patients were inoperable, as well as the fact that some patients were primarily operated on at other hospitals and referred to our hospital with recurrent disease. In conclusion, analysis of EHF mRNA expression in a large cohort of patients with primary OC, OC effusions, and MM effusions showed increased levels of this gene in OC effusions compared to MM effusions, supporting the use of the EHF mRNA assay as a novel diagnostic test in this
K. Brenne et al. setting. In addition, EHF mRNA levels were higher in primary OC compared to OC effusions. The biological rationale for this difference is yet to be established. Finally, higher EHF mRNA expression in OC effusions from patients with advanced disease correlated with poor survival, encouraging further research into the relationship between EHF mRNA levels and disease outcome in this cancer.
References [1] Oikawa T, Yamada T. Molecular biology of the Ets family of transcription factors. Gene 2003;303:11-34. [2] Oikawa T. ETS transcription factors: possible targets for cancer therapy. Cancer Sci 2004;95:626-33. [3] Seth A, Watson DK. ETS transcription factors and their emerging roles in human cancer. Eur J Cancer 2005;41:2462-78. [4] Hsu T, Trojanowska M, Watson DK. Ets proteins in biological control and cancer. J Cell Biochem 2004;91:896-903. [5] Sementchenko VI, Watson DK. Ets target genes: past, present and future. Oncogene 2000;19:6533-48. [6] Dittmer J. The biology of the Ets 1 proto-oncogene. Mol Cancer 2003;2:29. [7] Laudet V, Hanni C, Stehelin D, et al. Molecular phylogeny of the ETS gene family. Oncogene 1999;18:1351-9. [8] Kas K, Finger E, Grall F. ESE-3, a novel member of an epitheliumspecific Ets transcription factor subfamily, demonstrates different target gene specificity from ESE-1. J Biol Chem 2000;275:2986-98. [9] American Cancer Society. Cancer facts and figures 2010. Estimated new cancers cases and deaths by sex for all sites. Available at: http:// www.cancer.org/Research/CancerFactsFigures/CancerFactsFigures/ cancer-facts-and-figures-2010 (Accessed July 31, 2010). [10] Davidson B. Malignant effusions: from diagnosis to biology. Diagn Cytopathol 2004;31:246-54. [11] Davidson B. Biological characteristics of cancers involving the serosal cavities. Crit Rev Oncog 2007;13:189-227. [12] Davidson B, Zhang Z, Kleinberg L, et al. Gene expression signatures differentiate ovarian/peritoneal serous carcinoma from diffuse malignant peritoneal mesothelioma. Clin Cancer Res 2006;12:5944-50. [13] Beillard E, Pallisgaard N, van der Velden VH, et al. Evaluation of candidate control genes for diagnosis and residual disease detection in leukemic patients using “real-time” quantitative reverse-transcriptase polymerase reaction (RQ-PCR)—a Europe against cancer program. Leukemia 2003;17:2474-86. [14] Marchevsky AM. Application of immunohistochemistry to the diagnosis of malignant mesothelioma. Arch Pathol Lad Med 2008; 132:397-401. [15] Ordóñez NG. Value of immunohistochemistry in distinguishing peritoneal mesothelioma from serous carcinoma of the ovary and peritoneum: a review and update. Adv Anat Pathol 2006;13:16-25. [16] Ordóñez NG. What are the current best immunohistochemical markers for the diagnosis of epithelioid mesothelioma? A review and update. HUM PATHOL 2007;38:1-16. [17] Suster S, Moran CA. Applications and limitations of immunohistochemistry in the diagnosis of malignant mesothelioma. Adv Anat Pathol 2006;13:316-29. [18] Fetsch PA, Abati A. Immunocytochemistry in effusion cytology. Cancer 2001;93:293-308. [19] Davidson B, Nielsen S, Christensen J, et al. The role of desmin and Ncadherin in effusion cytology. Am J Surg Pathol 2001;25:1405-12. [20] Bassarova AV, Nesland JM, Davidson B. D2-40 is not a specific marker for cells of mesothelial origin in serous effusions. Am J Surg Pathol 2006;30:878-82. [21] He J, Pan Y, Hu J, et al. Profile of ETS gene expression in human breast carcinoma. Cancer Biol Ther 2007;6:76-82.
EHF in serous ovarian carcinoma [22] Galang CK, Mullers WJ, Foos G, et al. Changes in the expression of many ETS family transcription factors and of potential target genes in normal mammary tissue and tumors. J Biol Chem 2004;19:11281-92. [23] Hufton SE, Moerkerk PT, Brandtwijk R, et al. A profile of differentially expressed genes in primary colorectal cancer using suppression subtractive hybridization. FEBS Lett 1999;463:77-82. [24] Cangemi R, Mensah A, Albertini V, et al. Reduced expression and tumor suppressor function of the ETS transcription factor ESE-3 in prostate cancer. Oncogene 2008;27:2877-85. [25] Gonzalez HE, Gujrati M, Frederick M, et al. Identification of 9 genes differentially expressed in head and neck squamous cell carcinoma. Arch Otolaryngol Head Neck Surg 2003;129:754-9. [26] Yuan Y, Nymoen DA, Dong HP, et al. Expression of the folate receptor genes FOLR1 and FOLR3 differentiates ovarian carcinoma from breast carcinoma and malignant mesothelioma in serous effusions. HUM PATHOL 2009;40:1453-60. [27] Yuan Y, Nymoen DA, Stavnes HT, et al. Tenascin-X is a novel diagnostic marker of malignant mesothelioma. Am J Surg Pathol 2009;33:1673-82. [28] Davidson B. Anatomic site-related expression of cancer-associated molecules in ovarian carcinoma. Curr Cancer Drug Targets 2007;7: 109-20.
505 [29] Davidson B, Espina V, Steinberg AM, et al. Proteomic analysis of malignant ovarian cancer effusions as a tool for biologic and prognostic profiling. Clin Cancer Res 2006;12:791-9. [30] Cannistra SA. Cancer of the ovary. N Engl J Med 2004;351:2519-29. [31] Aletti GD, Gallenberg MD, Cliby WA, et al. Current management strategies for ovarian cancer. Mayo Clin Proc 2007;82:751-70. [32] Hoskins WI, McGuire WP, Brady MF, et al. The effect of diameter of largest residual disease on survival after primary cytoreductive surgery in patients with suboptimal residual epithelial ovarian carcinoma. Am J Obstet Gynecol 1994;170:974-9. [33] Rose PG, Nerenstone S, Brady MF, et al. Secondary surgical cytoreduction for advanced ovarian carcinoma. N Engl J Med 2004; 351:2489-546. [34] Davidson B, Reich R, Goldberg I, et al. Ets-1 messenger RNA expression is a novel marker of poor survival in ovarian carcinoma. Clin Cancer Res 2001;7:551-7. [35] Davidson B, Goldberg I, Gotlieb WH, et al. PEA3 is the second Ets family transcription factor involved in tumor progression in ovarian carcinoma. Clin Cancer Res 2003;9:1412-9. [36] Davidson B, Risberg B, Goldberg I, et al. Ets-1 mRNA expression in effusions of serous ovarian carcinoma patients is a marker of poor outcome. Am J Surg Pathol 2001;25:1493-500.