DNA methylation as a biomarker for the detection of hidden carcinoma in endometrial atypical hyperplasia

DNA methylation as a biomarker for the detection of hidden carcinoma in endometrial atypical hyperplasia

Gynecologic Oncology 135 (2014) 552–559 Contents lists available at ScienceDirect Gynecologic Oncology journal homepage: www.elsevier.com/locate/ygy...

2MB Sizes 4 Downloads 40 Views

Gynecologic Oncology 135 (2014) 552–559

Contents lists available at ScienceDirect

Gynecologic Oncology journal homepage: www.elsevier.com/locate/ygyno

DNA methylation as a biomarker for the detection of hidden carcinoma in endometrial atypical hyperplasia Hung-Cheng Lai a,b,c,d,e,f,1, Yu-Chi Wang f,1, Mu-Hsien Yu f, Rui-Lan Huang b, Chiou-Chung Yuan a,b, Kuan-Ju Chen g, Chia-Chun Wu h, Kai-Jo Chiang i, Tai-Kuang Chao c,g,⁎ a

Department of Obstetrics and Gynecology, School of Medicine, College of Medicine, Taipei Medical University, Taipei, Taiwan Department of Obestetrics and Gynecology, Shuang Ho Hospital, Taipei University, Taipei, Taiwan Department and Graduate Institute of Biochemistry, National Defense Medical Center, Taipei, Taiwan, ROC d Department of Clinical Pharmacology, Xiangya Hospital, Central South University, Changsha 410008, PR China e Institute of Clinical Pharmacology, Central South University; Hunan Key Laboratory of Pharmacogenetics, Changsha 410078, PR China f Department of Obstetrics and Gynecology, Tri-Service General Hospital, National Defense Medical Center, Taipei, Taiwan, ROC g Department of Pathology, Tri-Service General Hospital, National Defense Medical Center, Taipei, Taiwan, ROC h Department of Orthopaedic Surgery, Tri-Service General Hospital, National Defense Medical Center, Taipei, Taiwan, ROC i Department of Nursing, Tri-Service General Hospital, National Defense Medical Center, Taipei, Taiwan, ROC b c

H I G H L I G H T S • AJAP1, HS3ST2 and SOX1 methylation analysis is a potential method for detection of endometrial carcinoma hidden in atypical hyperplasia. • Testing the methylation status of candidate genes may assist in devising an adequate treatment strategy prior to major surgery. • AJAP1, HS3ST2 and SOX1 expression may help to differentiate premalignant endometrium and endometrial carcinoma.

a r t i c l e

i n f o

Article history: Received 23 July 2014 Accepted 19 October 2014 Available online 23 October 2014 Keywords: Endometrium Atypical hyperplasia Endometrial carcinoma Epigenetics Methylation

a b s t r a c t Objective. Women with atypical hyperplasia (AH) are often found to have endometrial carcinoma (EC) at hysterectomy. The purpose of this study was to evaluate whether the hypermethylation of specific genes found by methylomic approaches to the study of gynecologic cancers is a biomarker for EC in women with AH. Methods. We evaluated the methylation of AJAP1, HS3ST2, SOX1, and PTGDR from 61 AH patients undergoing hysterectomy. Endometrial biopsy samples were analyzed by bisulfite conversion and quantitative methylationspecific polymerase chain reaction. A methylation index was used to predict the presence of cancer. To confirm the silencing effects of DNA methylation, immunohistochemical analysis of AJAP1, HS3ST2, and SOX1 was performed using tissue microarray. Results. Fourteen (23%) patients had EC at hysterectomy. AJAP1, HS3ST2, and SOX1 were highly methylated in the EC patients' biopsy samples (p ≤ 0.023). AJAP1, HS3ST2, and SOX1 protein expression was significantly higher in patients with AH only (p ≤ 0.038). The predictive value of AJAP1, HS3ST2, and SOX1 methylation for EC was 0.81, 0.72, and 0.70, respectively. Combined testing of both AJAP1 and HS3ST2 methylation had a positive predictive value of 56%, methylation of any one of AJAP1, SOX1, or HS3ST2 had a 100% negative predictive value. Conclusions. Hypermethylation of AJAP1, HS3ST2, and SOX1 is predictive of EC in AH patients. Testing for methylation of these genes in endometrial biopsy samples may be a hysterectomy-sparing diagnostic tool. Validation of these new genes as biomarkers for AH screening in a larger population-based study is warranted. © 2014 Elsevier Inc. All rights reserved.


⁎ Corresponding author at: Department of Pathology, Tri-Service General Hospital, National Defense Medical Center, 3F, 325, Sec 2, Cheng-Gong Rd., Neihu district, Taipei City 114, Taiwan, ROC. Fax: +886 2 66000309. E-mail address: [email protected] (T.-K. Chao). 1 HC Lai and YC Wang contributed equally to this work.

http://dx.doi.org/10.1016/j.ygyno.2014.10.018 0090-8258/© 2014 Elsevier Inc. All rights reserved.

Endometrial carcinoma (EC) is one of the most common cancers of the female genital tract, although the incidence varies between countries [1]. Prolonged exposure to estrogen promotes the development of endometrial hyperplasia (EH), which leads to atypical hyperplasia (AH); 25–40% of patients with AH subsequently progress to EC. Although AH is the least common type of hyperplasia, it is the type

H.-C. Lai et al. / Gynecologic Oncology 135 (2014) 552–559

most likely to progress to type 1 EC, which accounts for more than 80% of uterine cancers [2]. EC is usually confined to the inner lining of the uterus, which can be removed by hysterectomy. Unfortunately, even in stage I EC, there is a significant risk of tumor recurrence, distant metastasis, and death [3]. EH is classified into two categories by the World Health Organization: 1) EH, including simple and complex hyperplasia without atypia, and 2) endometrial AH, including simple and complex hyperplasia with atypia [4]. Several studies have shown that cytological atypia, which is the major criterion for the diagnosis of AH and the most reliable indicator of progression from EH to EC, has poor reproducibility [2,5,6]. In 12.7–42.6% of cases, EC coexists in patients with a diagnosis of AH [7]. The high rate of unrecognized cancer among women diagnosed preoperatively with AH reflects the fact that the histologic criteria for differentiating AH from some types of EC on dilation and curettage (D&C) are controversial and subject to different interpretations [8–10]. Because of overlap in the histologic picture of AH and low-grade EC in the limited tissue sample evaluated before major surgery, differentiation on pathologic grounds alone can be very difficult or impossible [5, 10,11]. Although D&C before hysterectomy is the gold standard method for the diagnosis of endometrial lesions, detection of AH cannot rule out a more severe lesion [12,13]. EC found at the time of hysterectomy for AH may be associated with deep myometrial (10%) or cervical stroma (5%) invasion [14]. Hysterectomy is the main therapeutic modality for AH. Conservative approaches such as high-dose progestin may be acceptable treatment options in certain situations (e.g., to maintain fertility), but the risks of progression to malignancy and of concurrent EC remain high [14]. At present, there is no established biomarker to differentiate endometrial AH and EC. Such a marker could be hysterectomy sparing for AH patients without EC. Even when EC arising from endometrial premalignant lesions is clearly defined, the possibilities for EC screening are very limited. Reliable determination of the presence or absence of EC would allow for better surgical decisions about hysterectomy and staging. The reassurance of patients given fertility-sparing management for AH may alleviate unnecessary anxiety. There is, therefore, a need to develop new, molecular-based, complementary tools that could improve the pathological diagnosis. Epigenetic studies have demonstrated that silencing of genes, such as tumor-suppressor genes (TSGs), can act as a mechanism of carcinogenesis [15,16]. The addition of a methyl group to the cytosine–guanine (CpG) island results in gene silencing. Because epigenetic silencing of TSGs by promoter hypermethylation is observed commonly in human cancers, it is possible that DNA methylation could be used for the early diagnosis of cancer. This concept, and its application in gynecologic cancers, has been gaining acceptance during the past few years, especially in diagnosing and treating cervical cancer (screening and triage) and ovarian cancer (prognosis) [17–20]. However, similar studies of EC are relatively limited. It is known that the progression of EC involves a multistep process, and both genetic and epigenetic events have been shown to play important roles. Although gene promoter CpG islands epigenetically marked by de novo DNA methylation may serve as biomarkers in EC, they have been rarely studied in AH [21–24]. Such epigenetic biomarkers could be useful for identifying EC in AH. Our previous research on the epigenomics of cervical cancer using methylomic approaches identified several candidate genes that are methylated in cervical cancer tissues. Several candidate genes were significantly hypermethylated in CIN3+ lesions [19,25]. Because the uterine cervix and endometrium both originate from the Müllerian duct system, this close embryologic relationship between the uterine cervix and endometrium may be reflected in adulthood in the form of malignant lesions. We hypothesized that some of the genes hypermethylated in cervical cancer may also be hypermethylated in EC. We initially tested 28 development-related genes. To test further the feasibility of using these new biomarkers in identifying endometrial lesions, we converted the methylation analysis to a quantitative methylation-specific


polymerase chain reaction (QMSP) approach and tested its application value. We found that the following several genes were potentially implicated in endometrial carcinogenesis: adherens junction-associated protein 1 (AJAP1), heparan sulfate D-glucosamyl 3-O-sulfotransferase2 (HS3ST2), sex-determining region Y, box 1 (SOX1), prostanoid receptor gene, prostaglandin D2 receptor (PTGDR), and LIM-homeobox gene 1A (LMX1A). These candidate genes, which could be used for the triage of AH, were validated in The Cancer Genome Atlas (TCGA) EC database. The aim of the present study was to analyze the DNA methylation status of AJAP1, HS3ST2, SOX1, PTGDR, and LMX1A genes as biomarkers for EC diagnosis in patients with endometrial AH. The discovery of reliable epigenetic biomarkers for diagnosis may open a new avenue for the management of AH patients with and without EC. Materials and methods Patients and clinical samples Samples of endometrium from patients with endometrioid-type EC (n = 20; 8 G1 cases, 8 G2 cases, 4 G3 cases) and with dysfunctional uterine bleeding (n = 20) were included as cancer and normal controls, respectively. Specimens were obtained from tissue blocks for methylation analysis of the candidate genes. Endometrial biopsy tissues of patients with AH (n = 61) were collected for methylation analysis. All patients underwent hysterectomy within 3 months after endometrial sampling. The clinicopathologic characteristics of patients were recorded by the data managers from the Gynecologic and Pathological Center at the Tri-Service General Hospital, National Defense Medical Center, Taipei, Taiwan, R.O.C., who reviewed the patients' pathologic diagnosis and recorded their surgical status. The final diagnosis was made according to the worst pathologic finding from endometrial sampling or hysterectomy. Informed consent was obtained from all patients, and this study was approved by the Institutional Review Board of the Tri-Service General Hospital (TSGHIRB No: 2-101-05-041). DNA extraction, bisulfite conversion, and QMSP Before extraction of DNA from the paraffin-embedded tissue blocks, a 5 μm-thick section was cut from each tissue block and stained with hematoxylin and eosin (H&E) to confirm the histologic diagnosis and to define the purity of tumor or AH cells. For tissues in which the AH or EC area comprised ≥ 10% of the slide and the slide accounted for b20% of necrosis, the tissue sample was included in the DNA analysis. DNA was extracted from tissue samples using a commercial DNA extraction kit (QIAamp Tissue Kit; Qiagen, Hilden, Germany). DNA was prepared as described previously [26]. DNA from each tissue block was subjected to bisulfite methylation analysis. The DNA was treated with bisulfite using a CpGenome Universal DNA Modification Kit (Millipore, Bedford, MA) as described previously [26]. TaqMan-based QMSP (MethyLight) was performed after bisulfite treatment of denatured genomic DNA [27]. The methylation status of the candidate genes AJAP1, HS3ST2, SOX1, PTGDR, and LMX1A was tested. The primer sequences and cover promoter region of AJAP1, HS3ST2, and PTGDR are summarized in Table S1. The master mix and primers for SOX1 and LMX1A were purchased from iStat Biomedical Co. Ltd. The collagen type II α1 gene (COL2A) was used as an internal reference gene by amplifying non-CpG sequences. Each sample was analyzed in duplicate. In vitro Genome Universal Methylated Genomic DNA (Millipore) was used as a positive control because it is considered to represent 100% methylation of each gene. QMSP was performed in a total volume of 20 μL that contained 2 μL modified template DNA, 1 μL 20 × custom TaqMan reagent, and 10 μL LightCycler 480 Probes Master (Roche, Indianapolis, IN). The samples were subjected to an initial incubation at 95 °C for 10 min, followed by 50 cycles at 95 °C for 15 s, and annealing and extension for 1 min at the appropriate temperature, and then detected using the LightCycler 480 Real-Time PCR System


H.-C. Lai et al. / Gynecologic Oncology 135 (2014) 552–559

(Roche). The DNA methylation level was assessed as the methylation index, which was calculated as 10,000 × 2([Cp of COL2A] − [Cp of gene]) [28]. Failed amplification in QMSP was defined as a Cp value for COL2A N40. Tissue microarray and immunohistochemistry (IHC) Paraffin-embedded tissues for tissue microarray analysis, which were not paired to the tissue sections used for methylation analysis, were retrieved from the Department of Pathology, Tri-Service General Hospital, and tissue microarray slides were constructed according to a method published previously [29]. The tissue microarray comprised 100 samples, including 17 of normal endometrium, 16 of EH, 16 of AH, and 51 of endometrioid-type EC of stage T1a (n = 31), T1b (n = 13), T2 (n = 3), and T3 (n = 4). Tissue microarray sections were dewaxed in xylene, rehydrated in alcohol, and immersed in 3% hydrogen peroxide for 10 min to suppress endogenous peroxidase activity. Antigen retrieval was performed by heating each section at 100 °C for 30 min in 0.01 M sodium citrate buffer (pH 6.0). After three rinses (each for 5 min) in phosphate-buffered saline (PBS), sections were incubated for 1 h at room temperature with a mouse antibody specific for AJAP1 (1:100; Novus Biologicals, Littleton, CO), HS3ST2 (1:100; Abgent, San Diego, CA), or SOX1 (1:100; Chemicon, Temecula, CA) diluted in PBS. After three washes (each for 5 min) in PBS, sections were incubated with horseradish peroxidase-labeled immunoglobulin (Dako, Carpinteria, CA) for 1 h at room temperature. After three additional washes, peroxidase activity was visualized with a solution of diaminobenzidine at room temperature. To evaluate the immunoreactivity and histologic appearance, all tissue microarray slides were examined and scored concurrently by two of the authors. The immunoreactivity was graded arbitrarily and semiquantitatively by considering the percentage and intensity of staining over the whole section, as described previously [30,31]. Briefly, the intensity of staining for AJAP1, HS3ST2, and SOX1 in individual tumor cells was scored on a scale of 0 + (no staining), 1+ (weak intensity), 2 + (moderate intensity), and 3+ (strong intensity). Staining of cells was scored as negative if b30% of the slide area was positively stained, 1 + if 30–60% was stained, and 2+ if N 60% was stained. The absolute value of the percentage of cells at each intensity level was multiplied by the corresponding intensity value, and the products were added to obtain an immunostaining score of 0+, 1+, 2+, 3+, 4+, or 6+. As a negative control, the slide was treated similarly except for replacement of the primary antibody with nonimmune serum.

(p b 0.0001, p = 0.0002, p = 0.0004, p = 0.0035, respectively). The methylation index did not differ significantly between the nuclear grade G1, G2, and G3 cases (data not shown). These genes were chosen for further analysis. We searched for methylation data for EC from the TCGA database for validation. In the TCGA database, EC also had a high methylation level for AJAP1, HS3ST2, PTGDR, and SOX1 compared with normal controls (Fig. 2). LMX1A had a slightly elevated methylation index, but this did not differ significantly between patients with EC and normal controls (p = 0.5280; data not shown). Our finding for LMX1A was also consistent with the TCGA data (Fig. S1). In the TCGA analysis of overall survival, the presence of methylation of HS3ST2, SOX1, PTGDR, and LMX1A correlated significantly with survival. A high methylation index for HS3ST2, SOX1, PTGDR, and LMX1A was associated with a worse prognosis (Fig. S2). Methylation testing for identification of EC within endometrial AH We next identified 61 women with AH diagnosed by endometrial D&C, who had undergone hysterectomy within 3 months. In 14 of the 61 patients (23%), endometrioid-type EC was diagnosed at hysterectomy. In the remaining patients who were initially diagnosed with AH by endometrial D&C, subsequent hysterectomy revealed residual normal endometrium in 13 of 61 samples (21%), EH in 21 of 61 samples (35%), and AH in 13 of 61 samples (21%). These three groups were combined as a precancerous AH group. The median methylation index was significantly higher in the EC group than in the precancerous AH group for AJAP1, HS3ST2, and SOX1 (p = 0.0005, p = 0.014, p = 0.023, respectively), but there was no significant difference in methylation of PTGDR (p = 0.22; Figs. 3A and B). To assess the clinical application of this information, ROC curves were generated, and the area under the ROC curve (AUC) was calculated to discriminate between EC diagnosed at hysterectomy and the precancerous AH group (Fig. 3C). The best cutoff values for the methylation index for AJAP1, HS3ST2, and SOX1 were 94.58, 8.745, and 21.31, respectively. AJAP1 conferred the best accuracy. Sensitivity, specificity, AUC, positive predictive values (PPVs), and negative predictive values (NPVs) were also calculated. ROC curve analysis demonstrated that the sensitivity, specificity, accuracy, PPV, and NPV values were 86%, 72%, 0.81, 48%, and 94%, respectively, for AJAP1; 71%, 70%, 0.72, 42%, and 89%, respectively, for HS3ST2; and 71%, 60%, 0.70, 35%, and 88%, respectively, for SOX1 (Table 1). The performance of combined testing was also calculated (Tables 2A and B). Combined testing of both AJAP1 and HS3ST2 methylation had a PPV of 56% and an NPV of 89%. Combined testing for methylation of any one of AJAP1, SOX1, or HS3ST2 had a 100% NPV.

Statistical analysis The methylation values were analyzed with the two-sided Mann– Whitney U test to evaluate differences in methylation of the candidate genes between groups. Receiver-operating characteristic (ROC) curves were generated to confirm the accuracy of the diagnosis based on each candidate gene. Comparison of immunostaining scores between groups was performed using Student's t-test. p-Values b 0.05 were considered significant. SPSS for Windows (version 13; SPSS, Chicago, IL) was used for data entry and statistical analysis. Results DNA methylation of AJAP1, PTGDR, HS3ST2, SOX1, and LMX1A in EC We first examined the methylation status of the five genes AJAP1, PTGDR, HS3ST2, SOX1, and LMX1A from EC patients and compared it with that in samples from normal endometrium. We randomly chose 20 samples of EC and 20 normal controls for analysis. The methylation status of AJAP1, PTGDR, HS3ST2, and SOX1 in the EC and normal controls is shown in Fig. 1. The EC samples showed a higher methylation index for AJAP1, PTGDR, HS3ST2, and SOX1 compared with normal controls

Expression of AJAP1, HS3ST2, and SOX1 proteins in human precursor lesions but silencing in EC We evaluated AJAP1, HS3ST2, and SOX1 protein expression in endometrial lesions by performing IHC to analyze the endometrial lesion tissue arrays. We observed a wide variation in AJAP1, HS3ST2, and SOX1 expression between samples. Fig. 4A shows images of representative immunohistochemical staining for AJAP1, HS3ST2, and SOX1 in endometrial tissue samples of premalignant and malignant endometrial lesions, which reflect the incidence and intensity of immunoreactivity (Tables S2 to S7). A higher percentage of the premalignant endometrium samples with EH and AH expressed AJAP1 (15/16, 93.8%; 12/16, 75.0%), HS3ST2 (14/16, 87.5%; 12/16, 75.0%), and SOX1 (14/16, 87.5%; 11/16; 68.8%) compared with the EC samples (AJAP1, 26/51, 51.0%; HS3ST2, 26/51, 51.0%; and SOX1, 30/51, 58.8%; N 60% of the area was positive). The staining intensity for AJAP1, HS3ST2, and SOX1 was N2+ in all 32 of the premalignant endometrium samples with EH and AH (100%) compared with AJAP1 (32/51, 62.7%), HS3ST2 (32/51, 62.7%), and SOX1 (47/51, 92.2%) expression in the EC samples. The premalignant endometrium showed more intense immunostaining in the glandular epithelial cells compared with EC. Expression of AJAP1,

H.-C. Lai et al. / Gynecologic Oncology 135 (2014) 552–559


Fig. 1. Scatterplots showing the methylation index in EC and normal endometrium for AJAP1, PTGDR, HS3ST2, and SOX1 genes. The horizontal lines represent the median values. Em normal: normal endometrium; Em cancer: endometrial carcinoma.

Fig. 2. Comparison of the methylation levels of AJAP1, HS3ST2, PTGDR, and SOX1 in EC and normal controls in the TCGA endometrial cancer study.

556 H.-C. Lai et al. / Gynecologic Oncology 135 (2014) 552–559 Fig. 3. Characterization of methylation status in the precancerous AH and EC groups. Scatterplots showing the methylation index for AJAP1, PTGDR, HS3ST2, and SOX1 in samples diagnosed as AH in an endometrial D&C sample with a follow-up diagnosis at hysterectomy of normal endometrium, EH, AH, or EC (A). Combined group with a follow-up diagnosis of normal endometrium, EH, or AH (precancerous AH group) (B). The horizontal lines represent the median methylation values. (C) ROC curve analyses of the methylation of AJAP1, HS3ST2, SOX1, and PTGDR are shown. The AUC for each gene was calculated for the diagnosis of EC within AH. AH-normal: AH in D&C sample, residual normal endometrium on hysterectomy; AH-EH: AH in D&C sample, residual EH on hysterectomy; AH-AH: AH in D&C sample, residual AH on hysterectomy; AH-AC: AH in D&C sample, residual adenocarcinoma cells present on hysterectomy.

H.-C. Lai et al. / Gynecologic Oncology 135 (2014) 552–559 Table 1 The sensitivity, specificity, area under the receiver-operating characteristic curve (AUC), positive predictive value (PPV), and negative predictive value (NPV) of the potential markers of AJAP1, HS3ST2, and SOX1 genes DNA methylation to detect EC hidden in AH in endometrial D&C samples. Genes



AUC (95%CI)





86% (12/14) 71% (10/14) 71% (10/14)

72% (34/47) 70% (33/47) 60% (28/47)

0.81 (0.70, 0.93) 0.72 (0.56, 0.88) 0.70 (0.55, 0.86)

48% (12/25) 42% (10/24) 35% (10/29)

94% (34/36) 89% (33/37) 88% (28/32)



0.014 0.023

HS3ST2, and SOX1 was higher in AH than in EC (p = 0.019, p = 0.0043, and p = 0.038, respectively) (Fig. 4B). In addition, the more advanced T stages of EC had nonsignificantly lower immunostaining scores for AJAP1, HS3ST2, and SOX1. Discussion Endometrial carcinogenesis is a multistep process involving a precursor lesion with underlying genetic and epigenetic events. Therefore, molecular diagnostic methods have been proposed as new ancillary tools for the detection of undetected cancers and differential diagnosis of premalignant and malignant lesions. Abnormal patterns of DNA methylation have been recognized in various cancers. An increase in DNA methylation in gene promoter regions often precedes apparent malignant changes, which suggests that the assessment of DNA methylation could be used for the early diagnosis of cancer. Patients with AH are usually recommended to undergo hysterectomy because of the concern about undetected EC and the possible progression of AH to EC. Patients usually receive a simple hysterectomy for the pathologic diagnosis of endometrial AH. However, patients with invasive EC should undergo extended hysterectomy with or without lymph node dissection. Simple hysterectomy for invasive EC may inadvertently compromise surgical staging and jeopardize patients needing reoperation or radiotherapy, and those with tumor recurrence. Our results demonstrate the usefulness of assessing DNA methylation for the detection of hidden EC in AH. Patients with high methylation


levels of both AJAP1 and HS3ST2 should receive a thorough assessment, including a detailed endometrial biopsy or magnetic resonance imaging evaluation, before a major operation, or be checked by intraoperative frozen section because the PPV for EC could be as high as 56%. In addition, progestin therapy is commonly used for patients with AH who are reluctant to undergo hysterectomy, particularly because of fertility concerns. Daily treatment for about 3 months is recommended followed by another biopsy of the endometrium. AH persisting after 6 months of progestin therapy is strongly associated with treatment failure and should be an indication to pursue definitive surgical treatment in these patients [32]. Our results reveal that the NPV for undetected cancer can be 100% if there is no methylation of any one of AJAP1, HS3ST2, or SOX1. These epigenetic markers could provide another method for evaluating the risk of malignancy in women initially diagnosed with AH who want to preserve their fertility. Further longitudinal studies to evaluate the course of AH without DNA hypermethylation are warranted. The present study found different methylation levels of AJAP1, HS3ST2, and SOX1 in AH and EC. The protein expression, as shown by IHC, confirmed these differences. Based on the fact that it is difficult to make the differential diagnosis of AH and some ECs on morphologybased pathology alone, immunohistochemical analysis of these proteins, especially AJAP1, may help resolve the diagnoses. This benefit is additional to the prediction of EC in AH by DNA methylation. EH may be overdiagnosed when epithelial metaplastic changes occur in the context of simple or complex hyperplasia without atypia [10]. The possibility of overdiagnosis should be considered when the pathologic diagnosis is EC, but the tissue samples show high protein expression levels for AJAP1, HS3ST2, and SOX1. Our previous work identified SOX1 as a newly identified methylationsilenced gene in ovarian [20] and cervical [19,33] cancers. Epigenetic silencing of AJAP1 in glioblastoma has also been noted to be associated with shorter survival [34,35]. In addition, HS3ST2 could be used to segregate cervical high-grade squamous intraepithelial lesions/squamous cell carcinoma from normal/low-grade squamous intraepithelial lesions [36]. High levels of HS3ST2 methylation have also been shown in breast [37] and colon [38] cancers. PTGDR methylation is proposed as a biomarker for early detection of bladder cancer [39] and is associated with progression of neuroblastoma [40]. To our knowledge, there are no studies, validated by TCGA data, of these genes in EC. The present

Table 2 Combined two-gene (A) and three-gene (B) testing of DNA methylation to detect EC within AH in endometrial D&C samples. A. Genes Any one gene methylated


Sensitivity Specificity PPV NPV Both genes methylated Sensitivity Specificity PPV NPV SOX1 indicates sex-determining region Y, box 1. AJAP1, adherens junctions associated protein 1. HS3ST2, heparan sulfate (glucosamine) 3-O-sulfotransferase 2. PTGDR, prostaglandin D2 receptor.

93% 51% 36% 96% 64% 81% 50% 88%

AJAP1/HS3ST2 (13/14) (24/47) (13/36) (24/25) (9/14) (38/47) (9/18) (38/43)

93% 58% 39% 96% 64% 85% 56% 89%

SOX1/HS3ST2 (13/14) (27/47) (13/33) (27/28) (9/14) (40/47) (9/16) (40/45)

93% 50% 35% 96% 50% 19% 16% 56%

(13/14) (23/47) (13/37) (23/24) (7/14) (9/47) (7/45) (9/16)

B. Genes

Any one gene methylated Two of the three genes methylated All genes methylated

AJAP1/SOX1/HS3ST2 Sensitivity




100% (14/14) 79% (11/14) 50% (7/14)

43% (20/47) 72% (34/47) 87% (41/47)

34% (14/41) 46% (11/24) 54% (7/13)

100% (20/20) 92% (34/37) 85% (41/48)


H.-C. Lai et al. / Gynecologic Oncology 135 (2014) 552–559

Fig. 4. Examples of immunohistochemical staining and immunostaining scores for AJAP1, HS3ST2, and SOX1 in endometrial lesions. (A) H&E staining and immunohistochemical examination of endometrial lesions. H&E staining of histologically normal endometrium (A), EH (B), AH (C), and EC (D–F). Immunohistochemical examination of AJAP1 expression in normal endometrium (G), EH (H), AH (I), grade 1 EC (J), grade 2 EC (K), and grade 3 EC (L). Immunohistochemical examination of HS3ST2 in normal endometrium (M), EH (N), AH (O), grade 1 EC (P), grade 2 EC (Q), and grade 3 EC (R). Immunohistochemical examination of SOX1 in normal endometrium (S), EH (T), AH (U), grade 1 EC (V), grade 2 EC (W), and grade 3 EC (X). Magnification: ×200. (B) Semiquantitative comparison of AJAP1, HS3ST2, and SOX1 immunostaining scores between normal endometrium, EH, AH, and EC. *p b 0.05, AH vs EC.

H.-C. Lai et al. / Gynecologic Oncology 135 (2014) 552–559

study confirmed the important role of epigenetics in the development of EC and sheds light on the clinical applications of DNA methylation as a biomarker for the detection of hidden EC in AH. Despite these promising results, the current study has some limitations. Because of the small sample size, the cutoff value for each gene was defined from a hospital-based, retrospective, case-controlled study using a research platform that may not be directly applicable to the clinical setting or to wider populations. The application of these new biomarkers may provide a new molecular method of triage for these ambiguous endometrial sampling cases and warrants further validation in a larger population-based study. In addition, all patients in the current study were Asian. Although our results are consistent with the TCGA data, the extent to which the current results can be applied to other ethnic populations remains to be determined. In summary, the epigenetic hypermethylation of AJAP1, HS3ST2, and SOX1 genes has potential as a valuable biomarker for the identification of undetected EC within AH. Methylation silencing of AJAP1, HS3ST2, and SOX1 may represent a new biomarker for pathologic differential diagnosis of premalignant and malignant endometrial lesions. Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.ygyno.2014.10.018. Conflict of interest statement The authors declare that no conflict of interest exists.

Acknowledgments This study was supported by grants from the Tri-Service General Hospital (TSGH-C101-073, TSGH-C102-074, TSGH-C103-076, TSGHC103-006-008-S01, TSGH-C103-006-008-S02, and TSGH-C103-006008S03), NSC-102-2628-B-038-010-MY3, NSC-102-2314-B-016-038 and NSC-102-2314-B-016-043 from the National Science Council, Taiwan, Republic of China, TMUTOP103005-1 and 103TMU-SHH-11 from Taipei Medical University, MOHW103-TDU-B-212-113001 and MOHW103-TDU-PB-211-122025 from Ministry of Health and Welfare. References [1] Hsu YT, Gu F, Huang YW, Liu J, Ruan J, Huang RL, et al. Promoter hypomethylation of EpCAM-regulated bone morphogenetic protein gene family in recurrent endometrial cancer. Clin Cancer Res 2013;19:6272–85. [2] Kurman RJ, Kaminski PF, Norris HJ. The behavior of endometrial hyperplasia. A longterm study of “untreated” hyperplasia in 170 patients. Cancer 1985;56:403–12. [3] Mariani A, Webb MJ, Keeney GL, Lesnick TG, Podratz KC. Surgical stage I endometrial cancer: predictors of distant failure and death. Gynecol Oncol 2002;87:274–80. [4] Baak JP, Mutter GL. EIN and WHO94. J Clin Pathol 2005;58:1–6. [5] Mutter GL. Diagnosis of premalignant endometrial disease. J Clin Pathol 2002;55: 326–31. [6] Fu YS, Gambone JC, Berek JS. Pathophysiology and management of endometrial hyperplasia and carcinoma. West J Med 1990;153:50–61. [7] Litta P, Bartolucci C, Saccardi C, Codroma A, Fabris A, Borgato S, et al. Atypical endometrial lesions: hysteroscopic resection as an alternative to hysterectomy. Eur J Gynaecol Oncol 2013;34:51–3. [8] Allison KH, Reed SD, Voigt LF, Jordan CD, Newton KM, Garcia RL. Diagnosing endometrial hyperplasia: why is it so difficult to agree? Am J Surg Pathol 2008;32:691–8. [9] Kendall BS, Ronnett BM, Isacson C, Cho KR, Hedrick L, Diener-West M, et al. Reproducibility of the diagnosis of endometrial hyperplasia, atypical hyperplasia, and well-differentiated carcinoma. Am J Surg Pathol 1998;22:1012–9. [10] Silverberg SG. Problems in the differential diagnosis of endometrial hyperplasia and carcinoma. Mod Pathol 2000;13:309–27. [11] Mutter GL. Histopathology of genetically defined endometrial precancers. Int J Gynecol Pathol 2000;19:301–9. [12] Suh-Burgmann E, Hung YY, Armstrong MA. Complex atypical endometrial hyperplasia: the risk of unrecognized adenocarcinoma and value of preoperative dilation and curettage. Obstet Gynecol 2009;114:523–9. [13] Taylor PJ, Gomel V. Endometrial ablation: indications and preliminary diagnostic hysteroscopy. Baillieres Clin Obstet Gynaecol 1995;9:251–60.


[14] Rakha E, Wong SC, Soomro I, Chaudry Z, Sharma A, Deen S, et al. Clinical outcome of atypical endometrial hyperplasia diagnosed on an endometrial biopsy: institutional experience and review of literature. Am J Surg Pathol 2012;36:1683–90. [15] Agrawal S, Unterberg M, Koschmieder S, zur Stadt U, Brunnberg U, Verbeek W, et al. DNA methylation of tumor suppressor genes in clinical remission predicts the relapse risk in acute myeloid leukemia. Cancer Res 2007;67:1370–7. [16] Wu G, Guo Z, Chang X, Kim MS, Nagpal JK, Liu J, et al. LOXL1 and LOXL4 are epigenetically silenced and can inhibit ras/extracellular signal-regulated kinase signaling pathway in human bladder cancer. Cancer Res 2007;67:4123–9. [17] Lai HC, Lin YW, Huang RL, Chung MT, Wang HC, Liao YP, et al. Quantitative DNA methylation analysis detects cervical intraepithelial neoplasms type 3 and worse. Cancer 2010;116:4266–74. [18] Chao TK, Ke FY, Liao YP, Wang HC, Yu CP, Lai HC. Triage of cervical cytological diagnoses of atypical squamous cells by DNA methylation of paired boxed gene 1 (PAX1). Diagn Cytopathol 2013;41:41–6. [19] Lai HC, Lin YW, Huang TH, Yan P, Huang RL, Wang HC, et al. Identification of novel DNA methylation markers in cervical cancer. Int J Cancer 2008;123:161–7. [20] Su HY, Lai HC, Lin YW, Chou YC, Liu CY, Yu MH. An epigenetic marker panel for screening and prognostic prediction of ovarian cancer. Int J Cancer 2009;124: 387–93. [21] Di Domenico M, Santoro A, Ricciardi C, Iaccarino M, Iaccarino S, Freda M, et al. Epigenetic fingerprint in endometrial carcinogenesis: the hypothesis of a uterine field cancerization. Cancer Biol Ther 2011;12:447–57. [22] Arafa M, Somja J, Dehan P, Kridelka F, Goffin F, Boniver J, et al. Current concepts in the pathology and epigenetics of endometrial carcinoma. Pathology 2010;42:613–7. [23] Ignatov A, Bischoff J, Ignatov T, Schwarzenau C, Krebs T, Kuester D, et al. APC promoter hypermethylation is an early event in endometrial tumorigenesis. Cancer Sci 2010;101:321–7. [24] Jones A, Teschendorff AE, Li Q, Hayward JD, Kannan A, Mould T, et al. Role of DNA methylation and epigenetic silencing of HAND2 in endometrial cancer development. PLoS Med 2013;10:e1001551. [25] Chen YC, Huang RL, Huang YK, Liao YP, Su PH, Wang HC, et al. Methylomics analysis identifies epigenetically silenced genes and implies an activation of beta-catenin signaling in cervical cancer. Int J Cancer 2014;135:117–27. [26] Shih YL, Shyu RY, Hsieh CB, Lai HC, Liu KY, Chu TY, et al. Promoter methylation of the secreted frizzled-related protein 1 gene SFRP1 is frequent in hepatocellular carcinoma. Cancer 2006;107:579–90. [27] Ogino S, Kawasaki T, Brahmandam M, Cantor M, Kirkner GJ, Spiegelman D, et al. Precision and performance characteristics of bisulfite conversion and real-time PCR (MethyLight) for quantitative DNA methylation analysis. J Mol Diagn 2006;8: 209–17. [28] Huang RL, Chang CC, Su PH, Chen YC, Liao YP, Wang HC, et al. Methylomic analysis identifies frequent DNA methylation of zinc finger protein 582 (ZNF582) in cervical neoplasms. PLoS ONE 2012;7:e41060. [29] Hidalgo A, Pina P, Guerrero G, Lazos M, Salcedo M. A simple method for the construction of small format tissue arrays. J Clin Pathol 2003;56:144–6. [30] Jin JS, Hsieh DS, Loh SH, Chen A, Yao CW, Yen CY. Increasing expression of serine protease matriptase in ovarian tumors: tissue microarray analysis of immunostaining score with clinicopathological parameters. Mod Pathol 2006;19:447–52. [31] Sarmadi S, Izadi-Mood N, Sotoudeh K, Tavangar SM. Altered PTEN expression; a diagnostic marker for differentiating normal, hyperplastic and neoplastic endometrium. Diagn Pathol 2009;4:41. [32] Mentrikoski MJ, Shah AA, Hanley KZ, Atkins KA. Assessing endometrial hyperplasia and carcinoma treated with progestin therapy. Am J Clin Pathol 2012;138:524–34. [33] Chang CC, Huang RL, Wang HC, Liao YP, Yu MH, Lai HC. High methylation rate of LMX1A, NKX6-1, PAX1, PTPRR, SOX1, and ZNF582 genes in cervical adenocarcinoma. Int J Gynecol Cancer 2014;24:201–9. [34] Cogdell D, Chung W, Liu Y, McDonald JM, Aldape K, Issa JP, et al. Tumor-associated methylation of the putative tumor suppressor AJAP1 gene and association between decreased AJAP1 expression and shorter survival in patients with glioma. Chin J Cancer 2011;30:247–53. [35] Lin N, Di C, Bortoff K, Fu J, Truszkowski P, Killela P, et al. Deletion or epigenetic silencing of AJAP1 on 1p36 in glioblastoma. Mol Cancer Res 2012;10:208–17. [36] Lim EH, Ng SL, Li JL, Chang AR, Ng J, Ilancheran A, et al. Cervical dysplasia: assessing methylation status (Methylight) of CCNA1, DAPK1, HS3ST2, PAX1 and TFPI2 to improve diagnostic accuracy. Gynecol Oncol 2010;119:225–31. [37] Dietrich D, Lesche R, Tetzner R, Krispin M, Dietrich J, Haedicke W, et al. Analysis of DNA methylation of multiple genes in microdissected cells from formalin-fixed and paraffin-embedded tissues. J Histochem Cytochem 2009;57:477–89. [38] Gonzalo V, Lozano JJ, Munoz J, Balaguer F, Pellise M, Rodriguez de Miguel C, et al. Aberrant gene promoter methylation associated with sporadic multiple colorectal cancer. PLoS ONE 2010;5:e8777. [39] Reinert T, Modin C, Castano FM, Lamy P, Wojdacz TK, Hansen LL, et al. Comprehensive genome methylation analysis in bladder cancer: identification and validation of novel methylated genes and application of these as urinary tumor markers. Clin Cancer Res 2011;17:5582–92. [40] Sugino Y, Misawa A, Inoue J, Kitagawa M, Hosoi H, Sugimoto T, et al. Epigenetic silencing of prostaglandin E receptor 2 (PTGER2) is associated with progression of neuroblastomas. Oncogene 2007;26:7401–13.