Article Text

Gains of chromosome region 3q26 in intraepithelial neoplasia and invasive squamous cell carcinoma of the vulva are frequent and independent of HPV status
  1. S Aulmann1,
  2. J Schleibaum2,
  3. R Penzel1,
  4. P Schirmacher1,
  5. G Gebauer2,
  6. H P Sinn1
  1. 1Institute of Pathology, University of Heidelberg, Germany
  2. 2Department of Gynaecology and Obstetrics, University of Heidelberg, Germany
  1. Dr med Sebastian Aulmann, Department of Pathology, Im Neuenheimer Feld 220/221, 69120 Heidelberg, Germany; sebastian.aulmann{at}med.uni-heidelberg.de

Abstract

Aims: Two different forms of vulvar intraepithelial neoplasia (VIN) are recognised: (1) usual-type (bowenoid) VIN, which is related to high-risk papillomavirus infection, and (2) differentiated (simplex) VIN, which is associated with chronic inflammation. The aim of this study was to investigate the presence of chromosome 3q26 gains in the spectrum of precancerous lesions and invasive squamous cell carcinomas (SCCs) of the vulva.

Methods: 3q26 gains were analysed using fluorescence in situ hybridisation in a series of usual-type VINs, VINs of the differentiated type and invasive squamous cell carcinomas. In addition, all cases were examined for human papillomavirus (HPV) DNA, p53 mutations, and p16 and p53 protein expression.

Results: Gains of chromosome 3q26 were present in all VINs of the differentiated type and in 50% of the usual-type VIN lesions. 81% of SCCs were positive for 3q26 gains irrespective of the HPV status and of the associated precursor lesion. HPV-associated lesions exhibited the typical, strong cytoplasmic p16 accumulation while mutated p53 was only detected in HPV-negative VINs or SCCs, and was associated with an overexpression of p53 protein.

Conclusions: Immunohistochemical evaluation of p16 and p53 expression aids in the differential diagnosis of squamous cell alterations of the vulva. However, detection of 3q26 imbalance is of additional diagnostic value in difficult cases of HPV-unrelated usual-type VINs and VINs of the differentiated type.

Statistics from Altmetric.com

Take-home messages

  • Vulvar intraepithelial neoplasia can arise via human papillomavirus (HPV)-dependent and HPV-independent pathways.

  • The use of p16 is limited to HPV-dependent preneoplasia.

  • p53 Immunostaining is of additional value in p16-negative cases.

  • 3q26 gains occur in a high frequency regardless of HPV status.

  • Detection of 3q26 gains can aid in the differential diagnosis of vulvar precancerous lesions.

Squamous cell carcinomas (SCCs) of the vulva are relatively uncommon tumours that may affect women of any age. Depending on the prevalence of human papillomavirus (HPV), some cases arise via pathways mediated by the viral oncogenes E6 and E7.1 These tumours typically show a warty or basaloid histology and affect somewhat younger patients, often with a history of usual-type (Bowenoid) vulvar intraepithelial neoplasia (VIN). In contrast, keratinising SCCs frequently arise independently of HPV infections and are associated with differentiated (or simplex) VIN.2 Histologically, differentiated VINs are characterised by parakeratosis, thickened epidermis with elongated rete ridges and proliferation of abnormal keratinocytes in the parabasilar layer, often forming small nests with premature keratinisation within the rete ridges. These relatively subtle histological changes may be difficult to distinguish from simple hyperplasia in some cases. Especially in poorly oriented, small biopsy specimens, the diagnosis of preinvasive squamous lesions can be challenging. However, given the association with invasive SCC, the exact diagnosis of differentiated and usual-type VIN is of high clinical significance.

Gains of chromosome 3q are one of the most common chromosomal imbalances in SCCs, independent of the location. Using comparative genomic hybridisation, these alterations have been identified in a high proportion of SCCs of the head and neck,3 lung,4 oesophagus,5 cervix6 and vulva.7 As gains of the long arm of chromosome 3 are frequently accompanied by losses of chromosome 3p, isochromosome formation is believed to account for the majority of chromosome 3 imbalances. Possible candidate genes on chromosome 3q include hTERC,8 which encodes the RNA component of human telomerase, and PIK3CA, a regulator of cell growth and apoptosis that frequently contains mutations in different cancers.9

Figure 1 (A) Cytoplasmic accumulation of p16 in a case of human papillomavirus (HPV)-16 associated usual-type vulvar intraepithelial neoplasia (VIN). (B) p16-negative usual-type VIN. No HPV DNA was detected in this case. (C) Parabasilar extension of p53-positive nuclei in differentiated VIN. (D) Detection of 3q26 gains using two-colour fluorescence in situ hybridisation. In this case, a ratio of 2.05 3q26 signals (red) per cen7 signal (green) was observed.

Methods

A series of 62 preinvasive and invasive vulvar lesions, which had been diagnosed between 2002 and 2006, were retrieved from the archives of the Institute of Pathology, University of Heidelberg. All cases were re-evaluated independently by two pathologists (SA and HPS) who applied the current International Society for the Study of Vulvovaginal Diseases (ISSVD) classification.10 Invasive carcinomas were classified according to the World Health Organization (WHO).11 Immunohistochemical stains were performed with antibodies against p53 (1:100; clone DO-7; Novocastra, Newcastle, UK) and p16 (1:25; p16 research kit; DakoCytomation, Hamburg, Germany) using the avidin–biotin-complex method.

Fluorescence in situ hybridisation (FISH) with a 3q26/centromere 7 probe combination (DCS, Hamburg, Germany) was performed on paraffin-embedded tissue sections as described previously.12 Slides were viewed and photographed on an epifluorescence microscope (BX40; Olympus, Hamburg, Germany) and red (3q26) and green (centromere 7) hybridisation signals were counted in at least 50 nuclei. Signal ratios were calculated for each case by dividing the average 3q26 signal number by the average number of centromere 7 signals.

For PCR analyses, neoplastic tissue was microdissected using glass capillaries and digested using proteinase K (final concentration 0.5 μg/ml in 10 mM Tris/HCl, 1 mM EDTA, 0.2% Tween 20, pH 7.6). The presence of HPV DNA was evaluated using nested multiplex-PCR according to Sotlar et al.13 Exons 4–10 of the p53 gene were amplified using primers published by Done et al.14 Mutation screening was performed by single strand coded polymorphism electrophoresis as described previously.15 Briefly, samples were denatured using NaOH, run on 6.6% polyacrylamide gels, followed by silver staining. PCR products showing aberrant bands were sequenced on an ABI Prism 377 DNA sequencer (Applera, Darmstadt, Germany) using the DYEnamic ET terminator sequencing kit (GE Healthcare, Freiburg, Germany).

Results

There were 27 cases of usual-type VIN (according to the WHO classification, four had originally been classified as VIN 2 and 23 as VIN 3), five cases of differentiated VIN, and 28 invasive squamous cell carcinomas (19 keratinising, 6 non-keratinising, 2 basaloid and one warty SCC). Usual-type VIN occurred significantly earlier than VIN of the differentiated type (median age, 51 and 73 years respectively, p = 0.027, Wilcoxon rank test). For patients with SCC the median age at diagnosis was 55 years. The presence of high-risk HPV DNA as well as the immunohistochemical staining results are summarised in table 1 and in supplementary table 1.

Table 1 HPV subtypes, and p53 and p16 status

FISH with a 3q26/cen7 DNA probe combination was successfully performed in 60 samples. No specific hybridisation signals could be identified in three usual-type VIN and one SCC, probably because of poor fixation of the tissues. 3q26 gains with ratios ranging from 1.22 to 2.34 were identified in 12 usual-type VIN (50% of the 24 successfully hybridised cases, table 2). All of these cases represented relatively advanced lesions (VIN 3 according to the WHO classification). Usual-type VINs were further characterised by the frequent presence of high-risk HPV DNA. Using nested multiplex PCR, 20 of the cases (74%) contained high-risk HPV DNA (type 16 in 19 cases and type 18 in one case). All high-risk HPV-positive usual-type VINs were also strongly positive for cytoplasmic p16 and negative for p53. Two out of seven high-risk HPV-negative usual-type VINs showed p53 overexpression, and in both of these cases p53 DNA mutations were detected (table 3).

Table 2 3q26 Gains and 3q26/cen7 ratios
Table 3 p53 Mutations

In differentiated VIN, all five cases included in the study showed a gain in the chromosome 3q region, with between 1.33 and 2.78 3q26 signals per centromere 7 signal. All differentiated VIN were HPV negative and p16 negative. Nuclear and parabasal p53 staining of the dysplastic keratinocytes was observed in three of the five cases of differentiated VIN (60%), but no p53 mutation was detected.

In SCC, 22 of 27 cases showed increased 3q26/cen7 ratios (range, 1.23–2.87). In only four tumours (two basaloid, one keratinising and one warty SCC), high-risk HPV DNA was detected, but 22 cases (79%) displayed cytoplasmic p16 staining. While staining was intense and diffusely positive in 10 cases including the four high-risk HPV-associated tumours, the other 12 SCCs displayed only weak or moderate p16 labelling. In 10 SCCs without detectable HPV DNA, expression of p53 and p16 was detected, and p53 mutations were present in four SCCs (table 3).

Discussion

The development of vulvar SCC follows two different main pathways. In HPV-associated lesions, transcription of the viral E6 and E7 oncogenes, which can integrate into the cellular genome after an infection with high-risk papillomaviruses, leads to an increased resistance to apoptosis via p53 degradation as well as a deregulation of the cell cycle due to binding of E7 to the retinoblastoma protein.16 As a consequence of sustained E2F transcriptional activity, cytoplasmic accumulation of p16 is frequently observed in these lesions.17 Similar to previous studies,18 19 HPV 16 accounted for the majority of infections in our series of VIN and SCC, while HPV 18, 31 and 33 were detected in a small proportion of cases. Occasionally, co-infections with low-risk HPV 6/11 were observed. Of 27 usual-type VIN, high-risk HPV DNA was observed in 20 cases (74%). Immunohistochemically, these cases showed a strong cytoplasmic p16 staining and a lack of p53. However, only four of 28 vulvar SCCs (14%) contained high-risk HPV DNA. These tumours comprised two basaloid, one warty and one keratinising SCC. At the time of surgery, three of the patients were less than 45 years old. Similar to HPV-associated VIN, there was a complete lack of p53 staining and a cytoplasmic accumulation of p16.

However, most of the invasive SCCs appeared to have developed following a second, HPV-independent, pathway. In our series, these tumours affected somewhat older women compared with HPV-related carcinomas (median age, 56 years compared with 39 years). VINs of the differentiated type are believed to be a precursor of these tumours. Similar to previous reports,2 20 all five differentiated VINs in our series were HPV negative and developed in elderly women (median age, 73 years). Compared with the HPV-associated lesions, the biology of HPV-unrelated tumour development is far less well understood. Mutations of the p53 tumour suppressor gene occur in a subset of cases lacking the HPV-related E6 oncogene.21 In our series, two of seven HPV-negative usual-type VINs, as well as four SCCs, contained p53 mutations. Immunohistochemically, the cases with missense p53 mutations showed strong and homogenous overexpression of p53, possibly indicating an accumulation of non- or malfunctional protein. As expected, the two SCCs with nonsense mutations did not display any p53 staining. The remaining, HPV-unrelated SCCs and the five differentiated VINs showed varying p53 staining results, ranging from 0 to 90% positively stained nuclei. Interestingly, a number of HPV-negative SCC also displayed strong p16 staining. Since they lacked the E7 oncogene, this indicates a possible disruption of the retinoblastoma protein pathway via different, yet unknown, mechanisms.

Using FISH, we analysed gains of chromosome 3q26 in preinvasive and invasive squamous cell lesions of the vulva. In our series, all cases of differentiated VIN as well as 81% of the invasive SCC contained 3q26 gains. According to the ISSVD, the term usual-type VIN combines VIN 2 and VIN 3 lesions as defined in the WHO classification. We observed 3q26 gains in 57% of the usual-type VIN originally classified as VIN 3, while none of the four less-advanced VIN 2 lesions showed 3q26 imbalances. In a small series of cases, Stoltzfus et al. observed 3q26 gains in all six cases of vulvar SCC as well as in one case of vaginal intraepithelial neoplasia.22 In cervical dysplasia, increased 3q26 copy numbers have been reported in 76% of CIN 3, 63% of CIN 2 and in few cases of CIN 1/atypical squamous hyperplasia.8 As virtually all cases of CIN 3 or cervical SCC are HPV associated, p16 immunostaining has been shown to be a useful parameter in the differential diagnosis of these lesions. However, in vulvar squamous cell lesions, HPV-unrelated neoplastic transformation limits the use of p16 alone as a diagnostic aid. One useful additional marker is p53 that is overexpressed in a number of HPV-negative lesions. Especially in differentiated VIN, the parabasilar extension of p53-positive cells is a helpful diagnostic feature that we could confirm in three of five cases. Gains of 3q26 occur in HPV-related and unrelated preinvasive lesions and in a high percentage of SCCs making 3q26 copy number analysis using FISH an interesting additional parameter for the differential diagnosis of early vulvar neoplasia. Especially in differentiated VIN, which are sometimes difficult to distinguish from hyperplastic epithelial alterations, detection of 3q26 imbalances using FISH may be a valuable diagnostic tool.

In conclusion, gains of chromosome 3q26 frequently occur in usual-type VIN, differentiated VIN and SCC of the vulva. As the alteration can be seen in HPV-positive and -negative lesions, detection of 3q26 gains may be a useful adjunct to p16 and p53 immunohistochemistry in the differential diagnosis of vulvar preneoplasia.

REFERENCES

Supplementary materials

Footnotes

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.