Gynecologic Oncology 93 (2004) 435 – 440 www.elsevier.com/locate/ygyno Detection of hypermethylated genes in tumor and plasma of cervical cancer patients H.J. Yang, a V.W.S. Liu, a,* Y. Wang, a K.Y.K. Chan, a,b P.C.K. Tsang, a U.S. Khoo, b A.N.Y. Cheung, b and H.Y.S. Ngan a a Department of Obstetrics and Gynaecology, Faculty of Medicine, Queen Mary Hospital, The University of Hong Kong, Hong Kong SAR, China b Department of Pathology, Faculty of Medicine, Queen Mary Hospital, The University of Hong Kong, Hong Kong SAR, China Received 22 September 2003 Abstract Objective. The aim of this study is to investigate the prevalence of promotor CpG island methylation of the death-associated protein kinase (DAPK), p16, and O6-methylguanine-DNA methyltransferase (MGMT) genes in both tumor and plasma samples of cervical cancers. Methods. Methylation-specific PCR (MSP) was employed to detect promotor CpG island methylation of the DAPK, p16, and MGMT genes in 85 surgical tumor tissue samples and 40 pretreatment plasma samples from cervical cancers. Results. Promotor CpG island methylation of DAPK, p16, and MGMT was detectable, respectively, in 60%, 28.2%, and 18.8% of cases of cervical tumor DNA; and in 40%, 10%, and 7.5% of cases of patients’ plasma DNA. Moreover, at least one of the three methylated genes was detected in 75.3% (64/85) of cases of tumor and in 55% (22/40) of cases of plasma. Higher prevalence of methylation of DAPK was found in squamous cell carcinoma than in adenocarcinoma in both univariate and multivariate analysis. Methylation of p16 was significantly associated with that of MGMT in both univariate and multivariate analysis. The methylation pattern in primary tumor and plasma was found to be concordant in 23 patients with matched tissue and plasma samples. In cases positive for DAPK and p16 methylation in tumor, detection in the paired plasma sample was 64.3% (9/14) and 33.3% (3/9), respectively. Conclusions. Promotor CpG island methylation is a frequent event in cervical carcinogenesis. Detection of the methylated sequences in the circulation suggests that plasma DNA methylation warrants further study to determine its potential role in cancer management. D 2004 Elsevier Inc. All rights reserved. Keywords: Cervical cancer; Promoter CpG island hypermethylation; DAPK; p16; MGMT Introduction Cervical cancer is the second most common cancer in women worldwide . Human papillomavirus (HPV) is the major causative agent [2– 5]. However, majority of patients with HPV-associated lesions such as cervical intraepithelial neoplasm (CIN) will remain stable or spontaneously regress over time . Therefore, other genetic and epigenetic events are likely to be involved in cervical carcinogenesis. Promoter CpG island hypermethylation is a frequent epigenetic event in many human cancers [7,8]. It is a potential pathway for tumor suppressor gene inactivation. * Corresponding author. Department of Obstetrics and Gynaecology, The University of Hong Kong, Room 747, Faculty of Medicine Building, 21 Sassoon Road, Hong Kong SAR, China. Fax: +852-2816-1947. E-mail address: [email protected] (V.W.S. Liu). 0090-8258/$ - see front matter D 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.ygyno.2004.01.039 The profile of gene hypermethylation is different for each cancer type. Some genes are hypermethylated in multiple tumor types, e.g., p16; while other changes are tumor-type specific, e.g., hypermethylation of p73 is only observed in hematological cancers . Promoter hypermethylation of the death-associated protein kinase (DAPK ), p16, and O6methylguanine-DNA methyltransferase (MGMT) genes has been reported in 30– 60% of cervical cancers, but found to be absent in normal tissues [10 – 13]. Furthermore, hypermethylation of p16 and MGMT seems to be an early event in cervical tumorigenesis [11,12]. Recently, hypermethylated genes have also been detected in DNA from plasma, serum, urine, and sputum by methylation-specific PCR (MSP) [14 –19]. As this method can detect as little as one methylated gene per 1000 unmethylated copies, it has sufficient sensitivity to detect low concentrations of tumor DNA in plasma. Compared to the large number of different mutations, promoter hypermethy- 436 H.J. Yang et al. / Gynecologic Oncology 93 (2004) 435–440 lation occurs over the same regions of a given gene in each form of cancer. Unlike allelic loss, detection of methylation constitutes a positive signal, which is easier to be detected against a background of normal DNA. Identification of the occurrence of methylation for a relatively small subset of genes in plasma might provide a potentially powerful system of biomarkers for cancer detection. Therefore, in this study, we examined the prevalence of promoter CpG island methylation of DAPK, p16, and MGMT in tissues and plasma of cervical cancer patients. Materials and methods Patients and tissue samples This study was approved by the Ethics Committee of the University of Hong Kong (EC 2011-03). One hundred and two cervical cancer patients, who had been diagnosed and treated in Queen Mary Hospital, Hong Kong, from January 1990 to February 2003, were recruited for this study. Tumor tissues were obtained from 62 patients. Pretreatment peripheral blood samples were obtained from 17 patients. Paired tumor tissue and blood samples were available in another 23 patients. In addition, 100 normal tissue samples including white blood cells, cervix, and ovary from these patients were used as controls. Another 30 plasma samples taken from normal individuals were also used as controls for plasma DNA analysis. Informed consent was obtained from all individuals for the collection of tissue or blood. The histological type and grade of tumor were classified according to WHO criteria. The stage of each cancer was established according to the International Federation of Gynaecology and Obstetrics (FIGO) criteria. DNA isolation Nine-milliliter pretreatment peripheral blood, collected in a tube containing 1 ml of 3 M sodium citrate, was fraction- ated by 1/1 volume Ficoll and centrifuged at 2000 g for 20 min at room temperature. The top plasma layer was transferred to a microcentrifuge tube and centrifuged at 8000 g for 3 min to spin down all suspended cells and/ or cell debris. Then plasma was collected and stored at 20jC until further analysis. Plasma DNA was extracted using the High Pure PCR Product Purification Kit (Roche Applied Science, Mannheim, Germany) following procedures according to the manufacturer. Briefly, 200 Al plasma was mixed thoroughly with 1 ml binding solution, and then centrifuged at 10,000 g for 1 min. After discarding the flow through, DNA was rinsed in washing solution twice, then finally eluted by 50 Al sterile water. Genomic DNA was extracted from tissues using the standard Proteinase K treatment followed by phenol/chloroform extraction. Methylation-specific PCR (MSP) Sodium bisulfite treatment of genomic and plasma DNA was performed before MSP. Genomic DNA (1 Ag) or 50 Al plasma DNA was denatured with 0.3 M NaOH for 10 min at 37jC, and then bisulfite-treated as described by Herman et al. . DNA was then purified by Wizard DNA Cleanup System (Promega, Madison, WI). DNA was treated again with NaOH at room temperature, neutralized by ammonium acetate, and finally ethanol-precipitated. After treatment, the unmethylated cytosine will be converted to uracil, whereas the methylated cytosine is unchanged. Uracil is recognized as thymine by Taq polymerase. The treated DNA was suspended in 30 Al of Tris buffer and subjected to MSP using unmethylation-specific or methylation-specific primers. The three genes were analyzed individually. Primer sequence, annealing temperature, and the expected product size are listed in Table 1. Each 20 Al PCR contained 2 Al of bisulfite-treated genomic DNA or 5 Al of bisulfite-treated plasma DNA, 1 PCR buffer, 0.25 AM of each primer, 250 AM (each) dNTP mix, and 0.75 Unit of FastStart Taq DNA polymerase (Roche Applied Table 1 Primer sequences, annealing temperatures, and PCR product sizes Gene p16 MGMT DAPK Uf Ur Mf Mr Uf Ur Mf Mr Uf Ur Mf Mr Sequence Product size (bp) Annealing temperature (jC) 5V-TAA TTA GAG GGT GGG GTG GAT TGT-3V 5V-CAA CCC CAA ACC ACA ACC ATA A-3V 5V-TTA TTA GAG GGT GGG GCG GAT CGC-3V 5V-GAC CCC GAA CCG CGA CCG TAA-3V 5V-TTT GTG TTT TGA TGT TTG TAG GTT TTT GT-3V 5V-AAC TCC ACA CTC TTC CAA AAA CAA AAC A-3V 5V-TTT CGA CGT TCG TAG GTT TTC GC-3V 5V-GCA CTC TTC CGA AAA CGA AAC G-3V 5V-GGA GGA TAG TTG GAT TGA GTT AAT GTT-3V 5V-CAA ATC CCT CCC AAA CAC CAA-3V 5V-GGA TAG TCG GAT CGA GTT AAC GTC-3V 5V-CCC TCC CAA ACG CCG-3V 151 60 150 60 93 59 81 59 108 60 98 60 Uf : unmethylation forward primer; Ur: unmethylation reverse primer; Mf: methylation forward primer; Mr: methylation reverse primer. H.J. Yang et al. / Gynecologic Oncology 93 (2004) 435–440 Table 2 Prevalence of methylation of genes in tissue and plasma samples of cervical cancer patients Direct sequencing Methylated and unmethylated PCR products were confirmed by direct sequencing. The sequencing primers used were forward primers. Sequencing reactions were carried out using the BigDyek Terminator Cycle Sequencing V2.0 Ready Reaction Kit (Applied Biosystems, Foster City, CA) and analyzed automatically by ABI 310 DNA sequencer (Applied Biosystems). The cytosine in the unmethylated PCR products converts to uracil, whereas the cytosine in the methylated PCR products is unchanged. Methylated Methylated Methylated Overall DAPK p16 MGMT methylated Tumor (n = 85) Nontumor (n = 100) Plasma (n = 40) Normal plasma (n = 30) Paired cases Tissue (n = 23) Plasma (n = 23) 51 (60%) 0 24 (28.2%) 0 16 (18.8%) 0 64 (75.3%) 0 16 (40%) 0 4 (10%) 0 3 (7.5%) 0 22 (55%) 0 14 (60.8%) 9 (39.1%) 9 (39%) 3 (13%) 4 (17.4%) 1 (4.0%) 18 (78.3%) 12 (52.2%) 33.3% 25% 437 Statistical analysis Percentage of patients 64.3% with positive methylation in both plasma and tissues All the statistical analyses were performed with the software SPSS Windows version 10.0 (SPSS Inc., Chicago, IL). Numerical and categorical data were analyzed for statistical significance using t test and chi-square analysis, respectively. Parameters including age, histology type [squamous cell carcinoma (SCC) vs. adenocarcinoma (AC)], stage (early vs. late), grade of differentiation (G1 vs. G2 –3), and methylation status of the three genes were put to multivariate analyses using logistic regression. A P value of <0.05 was taken as significant. Science). Thermal cycling was initiated at 95jC for 5 min, followed by 45 cycles of 95jC for 30 s, the specific annealing temperature for 30 s, and extension temperature at 72jC for 30 s; and a final extension at 72jC for 2 min. A methylation-positive DNA control was made in vitro using SssI methylase (New England Biolabs, Beverly, MA) which methylated every cytosine of CpG dinucleotide in the DNA. An untreated blood DNA from a normal individual was used as negative control. The same PCR conditions were used for tumor, normal, and plasma DNA. PCR products were separated by 8% nondenaturing polyacrylamide gel. Results Prevalence of promoter CpG island hypermethylation in tissue samples of cervical cancer Methylation of the three genes were detected in cervical cancer samples but not in any of the normal tissue samples. The prevalence of methylation of the three genes in 85 Table 3 Prevalence of methylation of different genes in tissues of cervical cancer patients and relation to clinical data All subjects Mean age (years) P valuea Histological type Squamous Adenocarcinoma P valueb Stage Early stage Late stage P valueb Grade G1 G2 – G3 P valueb 85 50.2 85 61 24 85 41 44 83c 30 53 DAPK P16 MGMT Any gene + + + + 51 51.0 0.457 34 48.6 24 53.4 0.183 61 48.7 16 49.0 0.756 69 50.3 64 52.1 0.024 21 43.9 42 9 0.008 19 15 18 6 0.678 43 18 11 5 0.766 50 19 50 14 0.023 11 10 22 29 0.249 19 15 13 11 0.492 28 33 11 5 0.068 30 39 28 36 0.149 13 8 18 32 0.973 12 21 12 11 0.060 18 42 7 8 0.349 23 45 24 39 0.511 6 14 +: positive for methylation; : negative for methylation. a P value was obtained by t test. b P value was obtained by chi-square test. c Two patients were uninformative on the WHO grading. 438 H.J. Yang et al. / Gynecologic Oncology 93 (2004) 435–440 cervical cancers is summarized in Table 2. Methylation was detectable in at least one gene in 75.3% (64/85) of cervical cancers. Of these, 24.7% (21/85) of cases were positive for methylation in two genes, with another 4.7% (4/85) of patients were positive for methylation in all three genes. Methylation was more frequently detected in DAPK (60.0%; 51/85) than in p16 (28.2%; 24/85) and MGMT (18.8%; 16/ 85). Table 3 illustrates the distribution of methylation of the three genes according to clinicopathological parameters of age, histological type, stage, and grade. Patients with methylation were older than patients without methylation ( P = 0.024; Table 3) Methylation of the DAPK gene was found to be higher in SCC than AC by both univariate and multivariate analyses ( P = 0.008 and 0.012, respectively) (Table 3), The prevalence of DAPK methylation shows no significant difference between early stage and late stage ( P = 0.249); or between less differentiated and well-differentiated cancer ( P = 0.973). p16 or MGMT methylation was not associated with any clinical parameters. However, by both univariate and multivariate analyses, occurrence of methylation of the p16 gene was associated with the occurrence of methylation of the MGMT gene ( P = 0.011 and 0.047, respectively). Prevalence of promoter CpG island hypermethylation in plasma samples of cervical cancer The prevalence of methylation of DAPK, p16, and MGMT in plasma was 40.0% (16/40), 10.0% (4/40), and 7.5% (3/40), respectively. Methylated sequence of at least one gene was detected in a total of 55% (22/40) plasma samples (Table 2). In contrast, methylated DNA was not detectable in any of the 30 normal plasma samples. Methylation positivity was found in only one of the three genes for each case, with the exception of a patient with stage IB2 cancer in which methylation of both DAPK and p16 was detected in the plasma. A higher percentage of methylation of DAPK gene in plasma samples was detected than the other two genes (Table 2). Of the 40 plasma samples obtained, 23 of them had paired surgical tumor tissue available. We found a similar pattern of Fig. 1. Methylation-specific PCR analysis of DAPK hypermethylation. The control indicated the positions of the PCR products of the unmethylated allele (U) and the methylated allele (M). In Case 1, promoter methylation was not detected in tumor tissue or matched plasma sample. In Case 2, promoter methylation was detected in both tumor and plasma samples. methylation changes for the three genes in the plasma samples to their paired tumor tissues (Fig. 1; Table 2). Methylation of DAPK, p16, and MGMT was, respectively, detected in 60.8% (14/23), 39% (9/23), and 17.4% (4/23) of the tissue samples of these paired cases. Of the tissue methylation positive patients, methylation in plasma for DAPK, p16, and MGMT was detected in 64.3% (9/14), 33.3% (3/9), and 25% (1/4) respectively. None of the patients without methylation in cervical cancer tissues was found to have methylation in their plasma. The occurrence of methylation for the three genes in plasma was associated with that in their paired tumor tissues ( P = 0.014, Fisher’s Exact Test). Discussion Since methylation changes are sometimes associated with ageing of normal epithelium , only those methylation markers that are always unmethylated in normal cells would be potentially useful for cancer detection. Methylation of DAPK, p16, and MGMT has been reported in many cancers such as esophagus, lung, colon, and cervix. However, they have not been found methylated in normal tissues including buffy coat, bronchial brush, and urine samples from healthy subjects as well as from normal tissues of cancer patients [8,14,16,22,23]. Hence, in this study, we determined the prevalence of methylation of these three genes in both tissue and plasma samples from cervical cancers. Consistent with others, no methylation of the three genes was detected in normal tissues in the present study. In contrast, methylation of at least one of the three genes studied, DAPK, p16, and MGMT, was detected in about 75.0% of cervical cancers. A similar frequency (79%; 42/ 55) was demonstrated for the methylation of at least one of the p16, APC, HIC-1, DAPK, MGMT, and E-cadherin genes studied by Dong et al. , although additional three genes were included. In their study, the prevalence of methylation of DAPK, p16, and MGMT gene was 51%, 30%, and 8%, respectively. The highest prevalence of methylation among all genes was also observed in the DAPK, as demonstrated in this study. The overall findings suggest that methylation of DAPK, p16, and MGMT occurs frequently and may be used as markers for cancer detection. Dong et al.  found that promotor methylation of DAPK and p16 was more common in SCC cases than in AC cases. In this study, we also found that methylation of DAPK but not p16 was more common in SCC than AC. This indicates that methylation of particular genes may preferentially occur in particular cell types. Recently, Cohen et al.  demonstrated that RASSF1A promoter was hypermethylated in 45% (5/11) of AC, but not in any 31 cases of SCC. Thus, promoter hypermethylation may also be a cell-type-specific event. Great precautions should be taken when selecting methylated DNA markers for cancer detection. Promoter hypermethylation has been reported in the plasma/serum of various human cancer patients, including H.J. Yang et al. / Gynecologic Oncology 93 (2004) 435–440 colorectal cancer  and liver cancer . However, methylation in plasma/serum samples from cervical cancers has not yet been reported. In all 40 plasma samples studied, methylation of at least one gene was detected in about 55% of plasma. Furthermore, the occurrence of methylation for the three genes in plasma was associated with that in their paired tumor tissues. And no methylated sequence was detected in plasma from those negative for methylation in paired tissues, implying an absence of false-positive methylation in this study. Thus, methylated genes could also be detected in plasma DNA of cervical cancer patients. The presence of tumor-specific and methylated gene sequences in peripheral blood may be developed as a noninvasive approach for cancer detection. It is possible to detect abnormality using several frequently methylated markers in majority cases of cancers. Further analysis in a much larger sample size is warranted. Recently, Smiraglia et al.  demonstrated that DNA methylation was dynamic using restriction landmark genomic scanning. In paired primary and metastatic tumors from 13 patients of head and neck squamous cell carcinoma, the set of methylated loci found in primary tumors was found to be different from those detected in metastatic tumor. Some loci were common in both tumor; others were either detected only in the primary or metastatic tumor. This indicates that DNA methylation is a reversible process and may change at different stages of cancer development. Nuovo et al.  used MSP in situ hybridization to demonstrate that hypermethylation of p16 detected in the cancer lesion was also present in precursor lesions adjacent to cervical cancer. Virmani et al.  also reported the detection of p16 and MGMT methylation in CIN samples and they suggested that p16 and MGMT methylation were intermediate events in cervical carcinogenesis. Though in our study, no significant difference in methylation of p16 or MGMT was detected between disease stages or grades, the significant association between p16 and MGMT methylation suggested that these two genes are closely related during cervical carcinogenesis. Other genes such as RARb and GSPT1 have been proposed to occur early in cervical cancer, while FHIT methylation was a late, and tumor-associated event . This dynamic nature of methylation represents a significant difference between epigenetic and genetic changes. Thorough study of the prevalence of gene methylation at different stages of cancer should be carried out before methylated DNA markers could be used for cancer management. In conclusion, promotor CpG island methylation is a frequent event in cervical carcinogenesis. Demonstration of the methylated sequences in the circulation suggest that plasma DNA methylation warrants further study to determine its potential role in cancer management. Acknowledgments This study was supported by the University of Hong Kong Conference and Research Grant (10204392/11733/ 439 20900/323/01). The authors wish to acknowledge Daniel Fong for his advice in the statistical analysis of data. References  NIH. Cervical cancer. NIH Consensus Statement 1996;14:1 – 38.  Walboomers JM, Jacobs MV, Manos MM, Bosch FX, Kummer JA, Shah KV, et al. Human papillomavirus is a necessary cause of invasive cervical cancer worldwide. J Pathol 1999;189:12 – 9.  Schiffman MH, Castle P. Epidemiologic studies of a necessary causal risk factor: human papillomavirus infection and cervical neoplasia. J Natl Cancer Inst 2003;95:2E.  Bosch FX, Manos MM, Munoz N, Sherman M, Jansen AM, Peto J, et al. Prevalence of human papillomavirus in cervical cancer: a worldwide perspective. International Biological Study on Cervical Cancer (IBSCC) Study Group. J Natl Cancer Inst 1995;87:796 – 802.  Waggoner SE. Cervical cancer. Lancet 2003;361:2217 – 25.  Holowaty P, Miller AB, Rohan T, To T. Natural history of dysplasia of the uterine cervix. J Natl Cancer Inst 1999;91:252 – 8.  Esteller M. CpG island hypermethylation and tumor suppressor genes: a booming present, a brighter future. Oncogene 2002;21:5427 – 40.  Jones PA. DNA methylation and cancer. Oncogene 2002;21:5358 – 60.  Esteller M, Corn PG, Baylin SB, Herman JG. A gene hypermethylation profile of human cancer. Cancer Res 2001;61:3225 – 9.  Dong SM, Kim HS, Rha SH, Sidransky D. Promoter hypermethylation of multiple genes in carcinoma of the uterine cervix. Clin Cancer Res 2001;7:1982 – 6.  Virmani AK, Muller C, Rathi A, Zoechbauer-Mueller S, Mathis M, Gazdar AF. Aberrant methylation during cervical carcinogenesis. Clin Cancer Res 2001;7:584 – 9.  Nuovo GJ, Plaia TW, Belinsky SA, Baylin SB, Herman JG. In situ detection of the hypermethylation-induced inactivation of the p16 gene as an early event in oncogenesis. Proc Natl Acad Sci U S A 1999;96:12754 – 9.  Wong YF, Chung TK, Cheung TH, Nobori T, Yu AL, Yu J, et al. Methylation of p16INK4A in primary gynecologic malignancy. Cancer Lett 1999;136:231 – 5.  Soria JC, Rodriguez M, Liu DD, Lee JJ, Hong WK, Mao L. Aberrant promoter methylation of multiple genes in bronchial brush samples from former cigarette smokers. Cancer Res 2002;62:351 – 5.  Usadel H, Brabender J, Danenberg KD, Jeronimo C, Harden S, Engles J, et al. Quantitative adenomatous polyposis coli promoter methylation analysis in tumor tissue, serum, and plasma DNA of patients with lung cancer. Cancer Res 2002;62:371 – 5.  Zou HZ, Yu BM, Wang ZW, Sun JY, Cang H, Gao F, et al. Detection of aberrant p16 methylation in the serum of colorectal cancer patients. Clin Cancer Res 2002;8:188 – 91.  Hibi K, Taguchi M, Nakayama H, Takase T, Kasai Y, Ito K, et al. Molecular detection of p16 promoter methylation in the serum of patients with esophageal squamous cell carcinoma. Clin Cancer Res 2001;7:3135 – 8.  Goessl C, Krause H, Muller M, Heicappell R, Schrader M, Sachsinger J, et al. Fluorescent methylation-specific polymerase chain reaction for DNA-based detection of prostate cancer in bodily fluids. Cancer Res 2000;60:5941 – 5.  Silva JM, Dominguez G, Villanueva MJ, Gonzalez R, Garcia JM, Corbacho C, et al. Aberrant DNA methylation of the p16INK4a gene in plasma DNA of breast cancer patients. Br J Cancer 1999;80: 1262 – 4.  Herman JG, Graff JR, Myohanen S, Nelkin BD, Baylin SB. Methylation-specific PCR: a novel PCR assay for methylation status of CpG islands. Proc. Natl Acad Sci U S A 1996;93:9821 – 6.  Issa JP, Ahuja N, Toyota M, Bronner MP, Brentnall TA. Accelerated age-related CpG island methylation in ulcerative colitis. Cancer Res 2001;61:3573 – 7. 440 H.J. Yang et al. / Gynecologic Oncology 93 (2004) 435–440  Kwong J, Lo KW, To KF, Teo PM, Johnson PJ, Huang DP. Promoter hypermethylation of multiple genes in nasopharyngeal carcinoma. Clin Cancer Res 2002;8:131 – 7.  Maruyama R, Toyooka S, Toyooka KO, Harada K, Virmani AK, Zochbauer-Muller S, et al. Aberrant promoter methylation profile of bladder cancer and its relationship to clinicopathological features. Cancer Res 2001;61:8659 – 63.  Cohen Y, Singer G, Lavie O, Dong SM, Beller U, Sidransky D. The RASSF1A tumor suppressor gene is commonly inactivated in adenocarcinoma of the uterine cervix. Clin Cancer Res 2003;9: 2981 – 4.  Wong IH, Lo YM, Zhang J, Liew CT, Ng MH, Wong N, et al. Detection of aberrant p16 methylation in the plasma and serum of liver cancer patients. Cancer Res 1999;59:71 – 3.  Smiraglia DJ, Smith LT, Lang JC, Rush LJ, Dai Z, Schuller DE, et al. Differential targets of CpG island hypermethylation in primary and metastatic head and neck squamous cell carcinoma (HNSCC). J Med Genet 2003;40:25 – 33.
© Copyright 2016