Tel.: [+0034 848422186], Fax: +[0034 848422200]
Differential hypermethylation of genes in vulvar cancer and lichen sclerosus coexisting or not with vulvar cancer†
Article first published online: 1 DEC 2010
Copyright © 2010 UICC
International Journal of Cancer
Volume 128, Issue 12, pages 2853–2864, 15 June 2011
How to Cite
Guerrero, D., Guarch, R., Ojer, A., Casas, J. M., Méndez-Meca, C., Esteller, M., Barba-Ramos, E., Garcia-Bragado, F. and Puras, A. (2011), Differential hypermethylation of genes in vulvar cancer and lichen sclerosus coexisting or not with vulvar cancer. Int. J. Cancer, 128: 2853–2864. doi: 10.1002/ijc.25629
The authors declare that they have no competing interests
- Issue published online: 22 APR 2011
- Article first published online: 1 DEC 2010
- Accepted manuscript online: 23 AUG 2010 01:11PM EST
- Manuscript Accepted: 3 AUG 2010
- Manuscript Received: 15 APR 2010
- Department of Health of the Government of Navarra, Spain
- human papillomavirus;
- lichen sclerosus;
- vulvar cancer
- Top of page
- Material and Methods
- Supporting Information
Squamous cell carcinoma (SCC) of the vulva is a heterogeneous disease, associated or not with vulvar lichen sclerosus (LS). The precursor role of LS in vulvar cancer is unclear. We studied the epigenetic alterations of RASSF1A, RASSF2A, p16, TSP-1 and MGMT genes in vulvar SCCs, LS associated with SCC, isolated LS and normal vulvar skin. Gene hypermethylation and human papillomavirus presence were evaluated by methylation-specific PCR and PCR/reverse line blot, respectively. High-risk human papillomavirus types were present in 16.7% of the patients with vulvar SCC. There were increasing percentages of hypermethylation of genes from isolated LS to LS associated with vulvar SCC and vulvar SCC. The genes were hypermethylated more frequently in vulvar SCC associated with LS than in those not associated with LS, MGMT and RASSF2A being unmethylated in LS not associated with vulvar SCC. TSP-1 hypermethylation was related to recurrence in patients with vulvar cancer. Conclusions are as follows: (i) the epigenetic inactivation of genes is a common event in vulvar SCC and is also present in adjacent lesions, implying a possible precursor role for these alterations; (ii) MGMT and RASSF2A hypermethylation are present exclusively in vulvar SCC and LS associated with SCC, and absent from isolated LS; and (iii) TSP-1 hypermethylation is a bad prognosis factor in vulvar SCC.
Vulvar cancer (VC) accounts for 3–5% of all female genital cancers and is more frequent in women older than 75 years.1 Squamous cell carcinoma (SCC) is the most common type of VC and has been divided into two groups on the basis of the presence or absence of the human papillomavirus (HPV), which leads to different models of progression.2 It is well known that there are, on one hand, HPV-associated VCs of the warty or basaloid type and, on the other, keratinizing SCC, most of which are not HPV related.3
First, SCC associated with HPV, mostly HPV16, is the least frequent, accounting for one third of vulvar SCC. It is diagnosed at young ages and is associated with other types of cancer, such as vaginal and anal carcinoma. It shares risk factors with cervical cancer (multiple sexual partners, previous history of smoking and HPV presence and typical pathological features with warty or basaloid subtypes). The precursor lesion is of the classic vulvar intraepithelial neoplasia (VIN) type, and 4% of VIN have a risk of malignancy and frequently relapse spontaneously.4
On the other hand, SCC that is not associated with HPV is more frequent (66% of cases), typical of older ages and more frequently associated with lichen sclerosus (LS) or squamous cell hyperplasia.4, 5 The exact role of LS in pathogenesis of VC is not yet known, but the risk of developing vulvar SCC approaches 5% in women with these lesions.5 SCC is of the keratinizing type and is preceded by differentiated VIN, which is characterized by atypical basal keratinocytes with normal maturation. These precursor lesions are rarely diagnosed, their grade of malignancy is high and they exhibit immunohistochemical expression of altered p53 in 70% of cases.6 Several other pathological characteristics related to patient prognosis include tumor size and stage, lymph node status and depth of invasion.4, 5, 7
The molecular pathways leading to the development of the different types of SCC are not yet fully understood. In this regard, the gene inactivation caused by the hypermethylation of the promoter cytosine-phosphate-guanine (CpG) islands of tumor suppressor genes is a common event in cancer.8 There is a limited yet numerous set of candidate genes susceptible to hypermethylation that are involved in cell signalling, cell cycle control, repair of DNA lesions and/or neovascularization, e.g., RASSF1A, RASSF2A, p16, TSP-1 and MGMT.
The RAS association domain family (RASSF) 1A protein is a RAS effector that has important functions in cell cycle control, apoptosis, microtubule stabilization and motility. The RASSF1A gene is located at 3p21, a region frequently deleted or hypermethylated in breast, lung and cervical cancers.9 Another member of RASSF family, RASSF2A, is a negative effector of RAS protein and has been recently identified as a gene inactivated in colorectal and endometrial cancer.10
The p16 gene is crucial in cell cycle control because its encoding protein is an inhibitor of cyclin-dependent kinases 4 and 6. This gene is altered by several mechanisms in cancer (homozygous deletion, methylation of the promoter and point mutation), and its overexpression is associated with progression from benign to malignant lesions.11
Thrombospondin (TSP)-1 protein participates in tissue genesis and remodelling. It is considered to be a potent endogenous inhibitor of angiogenesis and, thus, useful for new antiangiogenic therapies.12 Its alteration has been described in cervical and penile cancers.13, 14
O(6)-methylguanine-DNA methyltransferase (MGMT) and other proteins, e.g., AlkB homologous proteins and base excision repair proteins, are key proteins for the repair of O(6)-methylguanine (O(6)MeG) lesions in DNA induced by common carcinogens and inflammation.15
The objectives of this study were to determine the methylation status of the RASSF1A, RASSF2A, p16, TSP-1 and MGMT gene promoters in vulvar SCC, to determine any correlations of these with clinicopathological variables and to analyze their presence in LS associated or not with vulvar SCC and vulvar normal skin adjoining these lesions.
Material and Methods
- Top of page
- Material and Methods
- Supporting Information
The group studied consisted of a series of 30 invasive vulvar SCC cases, 12 LS associated with 12 of the tumors mentioned previously, 12 normal vulvar skin adjoining to 12 of these LS and 21 LS not associated with vulvar SCC collected between January 2000 and March 2007 (Figs. 1a–1c). All these lesions and normal vulvar skin were examined carefully under the microscope by the pathologist to select more than 90% of representative cells using hematoxylin and eosin as a guide to obtain cell populations by scalpel. Normal vulvar skin did not display cytological signs of viral infection or cytological atypia (Supporting Information Fig. S1a). This study was approved by the local ethics committee. Clinical and pathological information from the patients was compiled in a central database.
The mean (range) ages of patients with vulvar SCC and LS not associated with vulvar SCC were 73.03 (37–86) and 57.1 (25–74) yr, respectively. All the patients with vulvar SCC were treated by primary surgical resection; 12 (40%) and 18 (60%) patients were treated by partial and by radical vulvectomy, respectively. None of the patients received radiation or chemotherapy before surgery.
The vulvar SCC tumors were staged according to the TNM classification4 (Table 1). Lymphadenectomy was performed on the basis of tumor stage, patient age and health status. Seventeen patients (56.7%) were treated by chemotherapy (5-fluorouracil + cisplatin), and 6 patients (20%) were treated with radiotherapy and chemotherapy according to standard protocols.16 The follow-up was generally by physical examination, pelvic computed tomography scan and radiography every 6 mo during the first 2 yr and annually thereafter in most cases. The mean values for follow-up, disease-free survival (DFS) and overall survival (OS) of the patients with VC were 15.4, 12.4 and 15.4 mo, respectively. During follow-up, 14 patients (46.7%) had a local recurrence and 1 patient (3.3%) had distant metastasis. Fourteen patients (46.7%) died from vulvar SCC and two patients (6.7%) died from causes other than cancer.
Data concerning the presence of LS associated with vulvar SCC were available for 22 patients (73.3%) of which the condition was present in 16 of these cases (72.7%). LS presence was inversely associated with HPV presence (p = 0.002). One of the 21 cases of LS not associated with vulvar SCC evolved to one of the vulvar SCC after 4 years.
DNA extraction and β-globin amplification
Four representative 5-μm thick tissue sections were deparaffinized in xylene and digested with proteinase K (10 mg/mL; Roche, Indianapolis, IN) in 50–400 μL buffer depending on the quantity of tissue, at 56°C overnight. After inactivation of proteinase K at 96°C for 10 min, 2 μL of the supernatant was used directly for PCR. To assess DNA quality, amplification of the β-globin gene was performed by PCR using primers BGPCO3 and BGPCO5 that generates a 209-bp product. All the types of lesion were analyzable; 4 of the 16 LS associated with vulvar SCC were analyzable due to tissue availability and/or good quality of DNA.
Detection and genotyping of HPV DNA
We used the PCR linked to a nonradioactive reverse line blot procedure as previously described.17 This technique amplifies a 150-bp fragment of viral DNA by adding GP5+/GP6+ biotinylated primers to the reaction mixture consisting of 5 μL of PCR buffer, 1.875 mM MgCl2, 0.25 mM deoxynucleotides, 25 pmoles of each primer and 1 unit of AmpliTaq DNA Polymerase (Applied Biosystems, Foster City, CA). The conditions used in the thermocycler were 4 min at 94°C, followed by 40 cycles of 1 min at 94°C, 2 min at 40°C and 1.5 min at 72°C and a final extension for 4 min at 72°C.
Thirty-six HPV genotypes were determined by hybridizing the GP5+/6+-PCR products with specific oligonucleotide probes containing a 5′-amino group on a carboxyl-coated nylon membrane (Biodyne C; Pall, East Hills, NY) in a miniblotter (MN45; Immunetics, Boston, MA). One case each positive for HPV16 and distilled water were included as positive and negative controls, respectively. A total of 37 probes were used to identify high-risk types (HPV16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 68, 73, 82/MM4 and 82/IS39), probable high-risk types (HPV26, 53 and 66) and low-risk types (HPV6, 11, 34, 40, 42, 43, 44, 54, 55, 57, 61, 70, 71/ CP8061, 72, 81/CP83104, 83/MM7, 84/MM8 and CP6108). These probes were kindly provided by Dr. X. Bosch (ICO, Spain). Hybridization was followed by incubation of the membrane with streptavidine-peroxidase conjugate (Roche, Indianapolis, IN) and enhanced chemiluminescence detection (GE Healthcare, Munchen, Germany).
Methylation status of the gene promoters was determined by methylation-specific PCR (MSP), a technique that takes advantage of chemical modification of unmethylated, but not methylated, cytosine to uracil. Specific primers used to amplify regions of interest (Sigma, St. Louis, MO) and annealing temperatures were: (i) for the RASSF1A gene, unmethylated reaction: 5′-TTT GGT TGG AGT GTG TTA ATG TG-3′ (forward primer) and 5′-CAA ACC CCA CAA ACT AAA AAC AA-3′ (reverse primer); methylated reaction: 5′-GTG TTA ACG CGT TGC GTA TC-3′ (forward primer) and 5′-AAC CCC GCG AAC TAA AAA CGA-3′ (reverse primer); (ii) for the RASSF2A gene, unmethylated reaction: 5′-AGT TTG TTG TTG TTT TTT AGG TGG-3′ (forward primer) and 5′-AAA AAA CCA ACA ACC CCC ACA-3′ (reverse primer); methylated reaction: 5′-GTT CGT CGT CGT TTT TTA GGC G-3′ (forward primer) and 5′-AAA AAC CAA CGA CCC CCG CG-3′ (reverse primer); (iii) for the p16 gene, unmethylated reaction: 5′-TTA TTA GAG GGT GGG GTG GAT TGT-3′ (forward primer) and 5′-CAA CCC CAA ACC CAC AAC CAT AA-3′ (reverse primer); methylated reaction: 5′-TTA TTA GAG GGT GGG GCG GAT CGC-3′ (forward primer) and 5′-GAC CCC CGA ACC GCG ACC CTA A-3′ (reverse primer); (iv) for the TSP-1 gene, unmethylated reaction: 5′-TTG AGT TTG TGT GGT GTA AGA GTA T-3′ (forward primer) and 5′-CCC CAC TAC CTA ACA CAC AAC T-3′ (reverse primer); methylated reaction: 5′GTT CGC GTG GCG TAA GAG TAC-3′ (forward primer) and 5′-CGC TAC CTA ACG CGC AAC T-3′ (reverse primer); (v) for the MGMT gene, unmethylated reaction: 5′-TTT GTG TTT TGA TGT TTG TAG GTT TTT GT-3′and 5′-AAC TCC ACA CTC TTC CAA AAA CAA AAC A-3; methylated reaction: 5′-TTT CGA CGT TCG TAG GTT TTC GC-3′and 5′-GCA CTC TTC CGA AAA CGA AAC G-3′. The annealing temperature for the unmethylated and methylated reactions was 62°C and 66°C for p16 and 58°C and 60°C for TSP-1, respectively. The annealing temperature for both methylated and unmethylated reactions was 60°C, 65°C and 67°C for RASSF1A, RASSF2A and MGMT genes, respectively. DNA from lymphocytes treated in vitro with SssI methylase (New England Biolabs, Ipswich, MA) was used as the positive control for methylated alleles of all the genes. DNA from normal lymphocytes was used as a negative control. The presence of PCR products was verified with 2% agarose gel electrophoresis, stained with ethidium bromide and examined under ultraviolet illumination.
Methylation-specific sequencing of CpG islands
To further validate the MSP results for RASSF1A, RASSF2A, p16, TSP-1 and MGMT genes, DNA from normal vulvar skin, isolated LS, LS associated with vulvar SCC and SCC tumors were bisulfite modified and the CpG islands located up- and/or downstream of the transcription initiation site of each gene was amplified by PCR. PCR products were gel purified and cloned into the pGEMT Easy Vector System (Promega) according to the manufacturer's protocol. Plasmid DNA from 10 randomly chosen independent clones was purified with the NucleoSpin Multi-96 Plus Plasmid (Macherey Nagel, Düren, Germany). Individual plasmids were then sequenced using the 3100 Genetic Analyzer (Applied Biosystems).
Tests were carried out on 4-μm thick paraffin-embedded sections of all lesions, which were deparaffinized and rehydrated. Antigen retrieval was performed by incubation in citrate buffer pH 6.0, at 95°C for 30 min. After blocking endogenous peroxidase with 3% hydrogen peroxide, slides were incubated with monoclonal antibodies against p16 (clone 16P07; prediluted; LabVision, Fremont, CA), TSP-1 (clone A6.1; 1:50; LabVision), MGMT (clone MT3.1; 1:40, LabVision) and p53 (clone D07-prediluted; DakoCytomation, Cambridgeshire, United Kingdom) for 25 min at room temperature. Expression was visualized using 3,3-diaminobenzidine tetrahydrochloride in 3% hydrogen peroxide and hematoxylin and eosin counterstain. Incubation in PBS instead of primary antibody constituted the negative control.
Immunohistochemistry (IHC) results were scored independently by two pathologists blinded to the diagnosis. They evaluated combined nuclear and cytoplasmic staining (p16), and separate cytoplasmic (TSP-1) or nuclear (MGMT and p53) staining.
IHC results were semiquantitatively scored according to previous criteria. The intensity of TSP-1 in the cytoplasm and p16 staining in cytoplasm and nucleus of tumor cells were scored as one of four categories: 0, negative (no positive cells); 1, weak (1–50% positive cells); 2, moderate (51–75%); 3, strong (76–100%), following previously established criteria.18, 19 The immunoreactivity of MGMT and p53 proteins was scored after counting 200 consecutive cells as 0 (no expression, <5% positive cells), 1+ (5–25% positive cells), 2+ (26–50%), and 3+ (>50%), as previously described.20, 21
The monoclonal CD31 antibody (Dako) at 1:30 dilution (incubation time: 45 min, room temperature) was used to analyze microvessel density (MVD). This value was determined as the mean counts of CD31-positive endothelial cell clusters present at 200× magnification in the three areas of the specimens with the highest number of vessels, as previously described.22 Large vessels with muscular walls or areas of fibrosis, necrosis and inflammation were excluded from the analysis. To evaluate results, the median value of the counts was taken as the threshold between low and high MVD. Results were scored independently by two pathologists with no previous information about the cases.
Molecular, pathological and clinical categorical variables were summarized as frequencies and percentages, and the distributions of molecular and pathological variables were compared by the χ2 test or Fisher's exact test. The survival analysis used to determine which variables appear to affect recurrence of the disease consisted of two steps. The first consisted of fitting Kaplan-Meier survival curves, for DFS and OS, considering each of the explanatory variables, and the log-rank test was used to evaluate differences among them. The second step included all the explanatory variables that had a trend toward significance (p < 0.1) in a multivariate Cox proportional hazard model. The level of statistical significance was set at p < 0.05.
- Top of page
- Material and Methods
- Supporting Information
HPV typing and molecular studies
All the lesions were positive for β-globin amplification and, hence, could be tested for HPV, except for 4 of 16 LS associated with SCC. HPV was present in 16.7% of the patients with vulvar SCC (five patients), containing high-risk types in all the positive samples. The subtypes were: HPV16 (one patient, 3.3% of the cases), HPV18 (one patient, 3.3%), HPV59 (two patients, 6.7%) and a combined infection low-high risk HPV6 and 59 in one patient (3.3% of the cases; Fig. 2). Warty and basaloid carcinomas (3 cases, 10%) were HPV positive for HPV16 and HPV18 (p = 0.002) and tended to be p16 positive, although the relationship did not quite reach statistical significance (p = 0.059). SCC tumors associated with LS were of keratinizing type. Vascular space invasion was noted in seven vulvar SCC tumors and was more frequent in patients with lymph node involvement at diagnosis (p = 0.061). We confirmed the presence of the HPV59 subtype in one LS associated with vulvar SCC present in its tumoral counterpart.
The lesions positive for HPV16 and HPV18 were present in warty lesions, whereas the lesion with combined infection was of basaloid type. The associations between HPV presence and pathological data are shown in Table 2.
On the other hand, HPV was present in 25% of the LS not associated with SCC (5 lesions), with the HPV16 as the most prevalent subtype, present alone (3 cases, 15% of the lesions), or in combination with low-risk subtype HPV6 (1 case, 5%). We found one patient with the high-risk HPV58 viral sequence. The signal of the virus was completely distinguishable from the low background signal (Supporting Information Fig. S2).
It was possible to analyze RASSF1A, RASSF2A, p16, TSP-1 and MGMT in all vulvar SCC, and they were found to be hypermethylated in 30.0, 60.0, 70.0, 36.7 and 36.7% of cases, respectively (Fig. 3). These proportions were higher in vulvar SCC associated with LS (71.4, 72.7, 78.6, 62.5 and 75%, respectively) than in those not associated with LS (28.6, 0.0%, 21.4, 37.5 and 25%, respectively; Supporting Information Table S1). It is remarkable that RASSF2A was exclusively methylated in vulvar SCC associated with LS (p = 0.026).
The p16 hypermethylation was clearly associated with lymph node involvement at the time of diagnosis of vulvar SCC (p = 0.013, Table 2), and the lack of expression of its coding p16 protein (p = 0.005; Fig. 1d). Curiously, one was positive for p16 by IHC and hypermethylated in p16. The accumulation of p16 protein occurred exclusively in patients who were HPV16 positive and HPV18 positive.
TSP-1 expression was scored intermediate-strong in 13 cases (43%) and negative-weak in 21% of vulvar SCC cases (56.7%; Fig. 1e). The pattern of expression of TSP-1 protein was heterogeneous in several cases that were also methylated (Patient 3, 5, 24, 25, 26 and 28; Supporting Information Fig. S1b), which showed different areas with strong and weak expression. Intermediate-strong TSP-1 expression was more frequent in late stage tumors (Stage III, 42.8%) than in Stages I and II (28.6% each). The negative-weak expression was not associated with TSP-1 hypermethylation, and the latter was associated with recurrence (p = 0.029) and vascular invasion (p = 0.041) in vulvar SCC.
Nuclear MGMT expression was absent in 12 cases (40%), 1+ in 8 (26.7%), 2+ in 7 (23.3%) and 3+ in 3 patients (10%; Fig. 1f). There was a clear association between MGMT hypermethylation and lack of expression (p < 0.001), because 10 of 12 cases negative for MGMT expression were methylated. Nevertheless, we did not find such an association between TSP-1 hypermethylation and TSP-1 expression.
The accumulation of nuclear p53 protein was associated with neoplastic changes. It was detected in vulvar SCC (90%), in LS associated with vulvar SCC (40%) and was negative for LS not associated with vulvar SCC and normal skin samples.
The microvessels detected by CD31 expression were analyzed in all cases; they had a median value of 19 (range: 6–62 microvessels/mm2). Half the tumors had a low MVD (≤19 microvessels/mm2; Fig. 1g), the rest had a high MVD (Fig. 1h). This character and nuclear p53 were not correlated with any of the other variables tested.
With respect to LS associated with vulvar SCC, RASSF1A, RASSF2A, p16, TSP-1 and MGMT were hypermethylated in 33.3, 8.3, 16.6, 50 and 41.7% of the lesions analyzed (12 lesions), respectively (Supporting Information Table S1; Table 3). These percentages were lower than those obtained for the vulvar SCC counterparts. In this regard, all the LS that were methylated had their counterpart for vulvar SCC methylated, except one case each for RASSF1A and p16 (Table 3). It is notable that the percentages of hypermethylation were higher in vulvar SCC than in associated lesions.
On the other hand, 52.4, 19 and 52.4% of LS not associated with vulvar SCC were methylated for RASSF1A, p16 and TSP-1 genes, respectively. RASSF2A and MGMT genes were not methylated in these lesions. Finally the normal skin samples analyzed in this study were unmethylated for RASSF1A, RASSF2A, p16 and MGMT genes except for the TSP-1 gene in 50% of the cases (6 patients). Nevertheless, despite the fact that the results of TSP-1 hypermethylation of normal vulvar skin and isolated LS were confirmed by MSP, the band intensity was much less than for the positive control and the LS associated with SCC and vulvar SCC (Supporting Information Fig. 3). The MSS of TSP-1 confirmed that there were very few methylated clones for normal skin and isolated LS, compared to LS associated with SCC and vulvar SCC (Supporting Information Fig. 4). The absence of hypermethylation of the other genes in normal vulvar skin and the increasing percentages of this alteration from isolated LS to LS associated with vulvar SCC and vulvar SCC associated with LS were also confirmed by MSS (Supporting Information Table S1 and Fig. 4).
The disease recurred more frequently in patients with lymph node involvement at the time of diagnosis (p = 0.043). Kaplan-Meier analysis showed that unfavorable histological grade, lymph node involvement and TSP-1 hypermethylation were associated with shorter DFS (p < 0.001, p = 0.043 and p = 0.002, respectively). Nevertheless, only unfavorable histological grade was a factor in poor OS (p < 0.001; Supporting Information Table S2). Patients with HPV-positive tumors had significantly shorter OS (p = 0.016) but similar DFS periods to those with HPV-negative tumors.
The multivariate Cox proportional hazard ratio analysis showed that TSP-1 hypermethylation was an independent prognostic factor for DFS independently of other significant factors as histological grade or lymph node involvement (hazard ratio: 5.41; p = 0.015). This also occurs with histological grade, lymph node involvement and HPV presence (Supporting Information Table S3).
- Top of page
- Material and Methods
- Supporting Information
Vulvar SCC is relatively infrequent, and its incidence has changed proportionally less than that of in situ VC.1 Nevertheless, the mortality and morbidity of these lesions are noticeable, because they are not frequently diagnosed until they have reached an advanced stage, as was the case in our study.23 The study of clinicopathological and molecular variables may yield additional information about the prognosis of the patients and the role of precursor lesions in this type of cancer.
The oncogenic role of HPV in the pathogenesis of malignant tumors is caused by viral E6 and E7 proteins. Thus, E6 protein of high-risk HPV targets p53 protein for proteasomal degradation. The negative expression of altered p53 and the lack of mutations of this gene seem to be more frequent in HPV-positive tumors,24 but in our report, there was not any association between altered expression of p53 and HPV. The proportion of cases with HPV in the present study was lower than previously reported with diverging prevalence rates.25 This is probably due to the many vulvar SCC associated with LS, which tend to be HPV negative.26 Hence, in our series, HPV infection and LS associated with SCC are mutually exclusive states except for one case, which supports the previous findings exclusively in differentiated VIN that is usually associated with LS.27
In the present study, we confirmed the association between HPV infection and warty and basaloid SCC types28 and the presence of HPV59 in 10% of our cases. HPV59 is an uncommon high-risk type related to cervical intraepithelial neoplasia and cervical cancer and has been found in higher frequencies in anal carcinoma and those of male genital organs. To our knowledge, the presence of this viral type in vulvar SCC has not been noted in any previous studies, including those with large populations.28
In our study, epigenetic alterations distributed equally in HPV-positive and HPV-negative vulvar SCC, providing possible evidence of the same role of this mechanism in both types of lesions.29 We also found in our study that HPV was a bad prognostic factor, although the number of HPV-positive patients was small. On the other hand, the histological grade was a marked prognostic factor for DFS and OS, along with other pathological variables, such as vascular invasion.30
To date, few reports have addressed the molecular alterations involved in vulvar SCC, associated or not with LS. The hypermethylation of RASSF1A, RASSF2A, MGMT, p16 and TSP-1 genes analyzed here has been described in cervical, ovarian and endometrial cancers, among others.31 To our knowledge, there are no previous data about the hypermethylation of RASSF2A, MGMT and TSP-1 genes in vulvar SCC and LS. The different involvement of the hypermethylation in vulvar SCC associated with LS compared to those not associated with LS found in our study is notable, whereby RASSF2A is a gene that is inactivated exclusively in vulvar SCC associated with LS; for this reason, it could be a valuable tumor-specific marker. Conversely, RASSF1A, a gene located at a region on 3p21, which is frequently deleted and hypermethylated in VIN and VC,32 was methylated in all types of lesions analyzed in this study (LS associated or not with SCC and SCC). The silencing of RASSF genes (RASSF1A and RASSF2A) will probably provide cell growth advantage, because of their suppressor functions as activation of apoptosis, cell cycle control and microtubule stabilization.
On the other hand, the proportion of methylated p16 cases in our study group is similar to that described previously.33 This alteration is associated with lymph node involvement at diagnosis, which is evidence of the value of this gene as a marker for the progression in this type of cancer as in other cancers, e.g., cervical cancer.34 Curiously, only HPV16 and HPV18 were associated with p16 overexpression,2 in contrast to high-risk HPV59 tumors, which were p16 negative.
In our study, TSP-1 hypermethylation was not associated with the lack of expression of the protein, because of its coexistence with a heterogeneous pattern of TSP-1 expression (areas with strong and weak expression) probably due to a clonal effect or additional mechanisms such as p53 regulation.35TSP-1 hypermethylation was distributed equally in HPV-positive and HPV-negative tumors, associated or not with LS, confirming previous reports.3 It was also present in normal vulvar skin, although with very low levels of methylation confirmed by MSS. On the other hand, intermediate-strong TSP-1 expression was more frequent in late stage tumors, in accordance with previous reports.36 This expression was reported to be associated with angiogenesis, a crucial mechanism in the progression to VC.37 However, we did not find such an association in our patients.
It is important to note that in our group, the epigenetic alteration of TSP-1 gene was associated with recurrence and vascular invasion. We have recently reported that this alteration is a frequent event in penile cancer associated directly with poor prognosis.13 Here, we confirm for the first time the similar role of this alteration in vulvar SCC, with respect to DFS. In addition, the proportion of surviving patients diagnosed with vulvar SCC with unmethylated TSP-1 promoter was higher than in their methylated counterparts.
MGMT hypermethylation, found exclusively in LS associated with SCC and vulvar SCC in our study, was described as a late event in oligodendrogliomas and lung carcinomas.38 The alteration is crucial because it is related to specific sensitivity to alkylating chemotherapy due to the role of this protein in the repair of DNA lesions. To our knowledge, it has not been previously described in vulvar SCC and LS associated with vulvar SCC. We demonstrate that the hypermethylation is the definitive alteration leading to the lack of expression of MGMT protein, similar to what occurs with p16.
With regards to LS lesions, we found that a proportion of LS not associated with vulvar SCC were HPV positive,39 and here, we describe for the first time the presence in LS of HPV58, a rare high-risk subtype prevalent in cervical cancers in eastern Asia40 and in head and neck tumors.41 The presence of HPV in LS was reported elsewhere,19 and the unexpected discovery of HPV in both LS and squamous hyperplasia was, to a considerable extent, independent of the presence or absence of atypia.
RASSF1A, p16 and TSP-1 hypermethylation were also present in variable proportions of LS included in this study, whereas RASSF2A and MGMT hypermethylation were restricted to LS associated with vulvar SCC. The early alteration of RASSF1A was found to be present in uterine cancer and premalignant prostatic epithelial tissue.42, 43 Moreover, p16 and TSP-1 hypermethylation were described as early biomarkers in lung and premalignant stages of gastric cancer, respectively, increasing in frequency with tumor progression.44 The p16 inactivation is considered to be an early event present in LS,33 and we found it in our cases with more involvement in vulvar SCC.
The presence of methylation in normal tissue was described elsewhere, e.g., in the case of RASSF1A gene in tissue surrounding renal carcinomas45 and TPEF and PAX6 genes in normal mucosa adjacent to bladder tumors.46 Nevertheless, this alteration is not present in normal vulvar skin adjoining to vulvar SCC, except for the case of TSP-1 hypermethylation. However, very few methylated clones were detected by sequencing.
On the other hand, RASSF2A and MGMT hypermethylation are not present in benign lesions but are late alterations, as occurs for MGMT in oral carcinomas.47 Consequently, these facts appear to confirm the late role of these genes and their value as specific tumor markers not only in vulvar SCC but also in their adjacent LS.
The oncogenic role of LS is not clear, and additional data regarding malignancy is crucial.5 Histologically diagnosed LS are reported to have a 3–5% prospective risk of malignant transformation to SCC.5 In our study, one of the isolated LS included evolved to vulvar SCC after 4 years, and this lesion shared the same methylation pattern as SCC except for the MGMT hypermethylation and p53 overexpression present only in the latter. The presence of coincident p53 mutations and p53 expression was described in nonneoplastic epithelial lesions of the vulva and their adjacent vulvar SCC, suggesting that these alterations may be intrinsic to the clonal evolution that leads to SCC.48 LS associated to tumor analyzed in this study shared the same methylation profile (RASSF1A, p16 and TSP-1) or tended to have slightly lower percentages than the corresponding vulvar SCC. Nevertheless, isolated LS and normal skin tended to have slightly less extensive or no methylation (RASSF2A and MGMT) than for vulvar and abnormal p53 expression SCC, except for TSP-1. This shows a marked increase in methylated genes from normal vulvar skin to premalignant lesions and VC, suggesting that CpG island hypermethylation may contribute to morphological transformation and malignant progression. Thus, epigenetic alterations have been linked to cancer progression,49 and the facts mentioned above lead us to hypothesize about the precursor oncogenic role of methylation in LS evolving to vulvar SCC. More studies with larger populations are needed to be able to address this matter.
In conclusion, we report here for the first time the hypermethylation of RASSF2A, TSP-1 and MGMT genes in vulvar lesions, the increasing percentages of RASSF1A, RASSF2A, p16, TSP-1 and MGMT hypermethylation from normal vulvar skin, isolated LS, LS associated with SCC and vulvar SCC and the role of TSP-1 gene in prognosis related to recurrence for patients with vulvar SCC. We also report the early involvement of hypermethylation in isolated LS (RASSF1A, p16 and TSP-1), the exclusive hypermethylation of MGMT and RASSF2A in LS associated with SCC and the higher percentages of gene hypermethylation in VC associated with LS, with RASSF2A as the only marker specific to this group of tumors, compared to vulvar SCC not associated with LS. Further studies will help to clarify the risk of LS evolving to vulvar SCC and the role of these alterations, which are crucial to the detection and management of patients with worse prognosis.
- Top of page
- Material and Methods
- Supporting Information
We are grateful to all the patients and the clinicians who participated in the study. We also thank Dr. A. Fernández and M. Berdasco from PEBC (IDIBELL, Barcelona, Spain) for their support with methylation-specific sequencing of RASSF1A, RASSF2A, p16, TSP-1 and MGMT genes; and B. Ibáñez, MC. Caballero, M. Murillo, L. Serrano and E. Reta from the Biomedical Research Center (Pamplona, Spain) and Dr B. Lloveras, J. Klaustermeier and M. Olivera from Institut Català d'Oncologia (ICO, Barcelona, Spain) for their technical assistance.
- Top of page
- Material and Methods
- Supporting Information
- 4Tumours of the vulva: epithelial tumors. In: TavasolliFA, DevileeP. World Health Organization classification of tumours. Pathology and genetics and tumours of the breast and female genital organs. Lyon, France: IARC Press, 2003. 316–25., .
- 43Comparison of DNA hypermethylation patterns in different types of uterine cancer: cervical squamous cell carcinoma, cervical adenocarcinoma and endometrial adenocarcinoma. Int J Cancer 2006; 118: 2168–71., , , , , , , , .
- Top of page
- Material and Methods
- Supporting Information
Additional Supporting Information may be found in the online version of this article.
|IJC_25629_sm_suppfig1.tif||3018K||Supporting Figure 1 Normal skin from a total vulvectomy analysed in our study, without any sign of lesion (Supplemental fig. 1a); heterogeneous pattern of TSP-1 expression, with positive and negative areas coexisting with hypermethylation of the gene (case 25) (Supplemental fig. 1b).|
|IJC_25629_sm_suppfig2.tif||1360K||Supporting Figure 2 Reverse line blot (RLB) of the LS not associated with vulvar SCC; the 37 types tested in RLB are listed in the Methods. High-risk type HPV16, HPV18, HPV58 and a combined infection of low-risk type HPV6 and high-risk type HPV59 were found.|
|IJC_25629_sm_suppfig3.tif||3063K||Supporting Figure 3 MS-PCR of TSP-1 gene in LS not associated with vulvar SCC. Both the unmethylated (U) and methylated (M) PCR products are shown for each case. The signal of the methylated band is less intense than the controls. DNA from normal lymphocytes (NL) and in vitro-methylated DNA (IVD) are used as a positive PCR control for the unmethylated and the methylated reactions, respectively; H2O: Negative control.|
|IJC_25629_sm_suppfig4.tif||7114K||Supporting Figure 4 Bisulphite-sequencing analysis of RASSF1A, RASSF2A, p16, TSP-1 and MGMT promoter methylation status in normal vulvar skin, isolated lichen sclerosus, and paired associated lichen sclerosus and vulvar tumor. For each clone the methylation status of analysed CpG sites is shown (open and filled circles represent unmethylated and methylated circles, respectively). Densely methylated clones are present in associated lichen sclerosus and vulvar tumours. Black arrows represent the location of the methylation specific PCR (MSP) primers. The location of each CpG site relative to the transcription initiation site is shown by a vertical bar.|
|IJC_25629_sm_supptable1.doc||43K||Supporting Table 1.|
|IJC_25629_sm_supptable2.doc||31K||Supporting Table 2.|
|IJC_25629_sm_supptable3.doc||39K||Supporting Table 3.|
Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.