|Home | About | Journals | Submit | Contact Us | Français|
Differentiated vulvar intraepithelial neoplasia (VIN) is presumed to be the precursor of invasive squamous cell carcinoma (SCC) of the vulva. It is commonly assumed that differentiated VIN is related to lichen sclerosus (LS). However, evidence for this is limited to a small number of studies describing epithelial alterations adjacent to vulvar SCC.
To study the histology and human papillomavirus (HPV) status in patients with a history of both LS and VIN without coexistent SCC.
Original biopsy specimens and surgical specimens of patients retrieved from the pathology files were revised for the presence of LS, VIN and (early) invasive SCC, specifically focused on the two different types of VIN: differentiated and undifferentiated. Thereafter, VIN lesions were tested for the presence of HPV DNA.
Twenty‐seven patients fulfilled the criteria for LS and VIN without SCC. In all 27 patients, LS was found to be related to undifferentiated VIN. Grading yielded the following results: VIN 1 (n=10), VIN 2 (n=11) and VIN 3 (n=6). Additionally, VIN lesions from 26 patients could be tested for the presence of HPV DNA. HPV DNA, predominantly type 16, was present in 8 (31%) of them. Seven of these eight patients had VIN 2 or 3. During follow‐up, three patients progressed to (early) invasive carcinoma. In two of these patients, differentiated VIN was observed overlying early invasive SCC.
VIN related to LS without coexisting SCC is likely to be undifferentiated, in contrast to what was previously thought. HPV DNA was demonstrated in 31% of the lesions, and was strongly related to high‐grade VIN.
Invasive squamous cell carcinoma (SCC) of the vulva often arises in association with other vulvar abnormalities.1,2,3,4,5,6,7,8,9 These abnormalities usually fall into two main categories, which can be considered as the main precursor states for invasive carcinoma of the vulva: vulvar intraepithelial neoplasia (VIN) and lichen sclerosus (LS).
According to the International Society for the Study of Vulvovaginal Diseases (ISSVD) and the International Society for Gynecological Pathologists, LS is a non‐neoplastic disorder of the vulvar skin and mucosa.10 Although the presence of LS in the adjacent skin of vulvar SCC is suggestive of a premalignant disease, longitudinal studies report only a slight tendency for LS to evolve into SCC (1–5%).11,12,13,14
VIN, on the other hand, is considered a pre‐neoplastic disorder of the vulvar skin,15 although progression to invasive carcinoma remains uncertain. Data on the follow‐up of untreated VIN 3 are scarce, and the natural history is mainly based on follow‐up after surgery. In the only systematic review on treatment of VIN, with data on 3322 patients, progression to invasive carcinoma was seen in 9% of the untreated patients and in 3% of patients after treatment.16 VIN can be classified into undifferentiated (classic or bowenoid) and differentiated (simplex) VIN.17 Undifferentiated VIN is associated with human papillomavirus (HPV), occurs predominantly in younger patients, and tends to be a multifocal and multicentric disease, whereas differentiated VIN is not related to HPV, is usually found in older women, and is commonly unifocal and unicentric. Differentiated VIN is rather uncommon. It is supposed to be associated with LS,17,18 although evidence for this is limited to a small number of studies describing epithelial alterations adjacent to vulvar SCC.4,7,19,20 Since differentiated VIN is often observed adjacent to or overlying superficially invasive SCC, it is presumed to be the precursor of most invasive SCCs of the vulva.18
Only four studies reported on the coexistence of LS and VIN, either differentiated or undifferentiated, without SCC.19,20,21,22 A major disadvantage of all the four studies is that the coexistence of LS and VIN was not the main research question, but was a coincident finding. In a series of 86 patients with LS, differentiated VIN was observed twice, as was undifferentiated VIN.20 Three other studies describing the histological features of VIN mentioned the presence of LS in 41 of 437 (9%) cases.19,21,22 HPV DNA testing was performed in only one of them.19 Since these three studies involved hardly any differentiated VIN, no conclusions can be drawn regarding the type of VIN and its relationship to LS.
As the relationship between differentiated or undifferentiated VIN and LS was never deliberately investigated, and since the role of HPV is not yet clarified, we studied the histology and HPV status in a large group of patients with a history of both LS and VIN without SCC.
All cases with both histologically diagnosed LS and VIN were retrieved from the pathology files (1984–2004) of the Academic Medical Centre, the VU University Medical Centre and the Netherlands Cancer Institute in Amsterdam, and the Erasmus University Medical Centre in Rotterdam, The Netherlands. Firstly, a computer search for lichen scherosus or sclerosis (et atrophicus) was performed (1807 specimens). Secondly, cases of women with anogenital LS were extracted from the list (1207 vulvar specimens). Thirdly, the pathology files of these patients with anogenital LS were searched for VIN without the initial presence of coexistent SCC (46 patients, 137 specimens).
Slides from the original biopsy specimens as well as those from all subsequent surgical specimens of these 46 patients were collected and revised by an experienced pathologist (FJWtK) for the presence of LS, VIN and (early) invasive SCC.
The diagnosis of LS was based on the presence of dermal hyalinisation, vacuolar alterations of the dermal–epidermal junction and a variable dense lymphocyte infiltrate, whether or not accompanied by epidermal atrophy, progressive loss of rete ridges, hyperplasia and/or acanthosis.23 VIN terminology was used according to the classification of the ISSVD.15 We specifically looked for the two different types of VIN: undifferentiated and differentiated. Undifferentiated VIN is characterised by disorientation and loss of squamous epithelial architecture and maturation, together with a variable degree of cellular atypia. Depending on the level of cellular disarray, undifferentiated VIN was graded into VIN 1, 2 or 3. Subsequently, all revised VIN 1 cases were stained with MIB 1, a cell proliferation marker. Differentiated VIN, on the other hand, shows little or no atypia above the basal or parabasal layers, and is therefore a far more subtle lesion than undifferentiated VIN. Enlarged prematurely differentiated keratinocytes with prominent eosinophilic cytoplasm and abnormal nuclei are found deep within the epidermis, frequently near the tips of elongated and branching rete ridges. Deeply located squamous whorls with or without keratin pearls are sometimes seen.15,17,24,25 Differentiated VIN should be regarded as VIN 3, owing to its supposed invasive potential.15,24
Depth of invasion, if present during follow‐up, is described as early invasive or invasive carcinoma.
All confirmed dysplastic lesions were tested for the presence of HPV. To this end, we extracted cellular DNA from corresponding formalin‐fixed, paraffin‐wax‐embedded tissue. Testing for HPV was conducted by using a standard GP5+/6+ PCR enzyme immunoassay, followed by reverse line blot analysis.26 This test is clinically validated.27 We used one assay for the 14 most prevalent high‐risk types of HPV (16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 66 and 68), and another for 22 low‐risk HPVs (6, 11, 26, 34, 40, 42, 43, 44, 53, 54, 55, 57, 61, 70, 71, 72, 73, 81, 82 (both subtypes MM4 and IS39), 83, 84 and CP6108). In addition, PCR amplification products were analysed for individual types of HPV (reverse line blot). A β‐globin test was performed as a control for the presence and quality of DNA in the paraffin‐wax‐embedded tissue.
In 31 of the 46 patients who met our computer search criteria, the diagnosis of both LS and VIN could be confirmed after histological revision. Fifteen patients did not fulfil the criteria for VIN (n=11), LS (n=2), or both VIN and LS (n=2). Four more patients presented with lesions suspected to be malignant at first onset of either LS or VIN, and were therefore excluded. Twenty‐seven patients (82 vulvar specimens) remained for evaluation of histology and HPV status. Their mean (range) age at first histological diagnosis was 64 (38–87) years.
In these 27 patients with both LS and VIN, VIN lesions were all classified as undifferentiated (100%). This was further graded into VIN 1 (n=10), VIN 2 (n=11) and VIN 3 (n=6; table 11).
To confirm the diagnosis, MIB staining in VIN 1 showed increased mitotic activity in all cases. A warty histological pattern was seen in 24 (89%) lesions; basaloid‐type VIN was found in one (4%) lesion, and in two (7%) VIN 1 lesions it was not possible to distinguish warty and basaloid VIN. We diagnosed LS and VIN at the same time in the same lesion in 13 patients (fig 11,, patients 1–3), mostly contiguous, and at the same time but in a different lesion in three more patients (fig 22,, patient 4). In the other 11 cases, a period of 1–9 years was observed between first diagnosis of the two skin disorders (fig 22,, patients 5 and 6). VIN arose in pre‐existent LS in nine of 11 (82%) cases, whereas LS was diagnosed in pre‐existent VIN twice.
Three patients developed (early) invasive SCC 1–4 years after the initial diagnosis of undifferentiated high‐grade VIN in pre‐existing LS. In one case the epithelium adjacent to invasive SCC showed undifferentiated VIN 3, whereas in the other two cases differentiated VIN was found in the epithelium surrounding early invasive SCC (fig 33,, patient 7).
Paraffin‐wax‐embedded tissue for HPV testing was available in 26 patients. HPV DNA was detected in eight of 26 (31%) tissue samples, whereas 18 lesions were found to be negative. Seven lesions contained HPV16 DNA, and one contained HPV59. There was a strong correlation between HPV‐positive lesions and high‐grade VIN. Whereas almost all VIN 1 lesions were HPV negative (89%), a considerable number of VIN 2 and 3 lesions were positive for HPV (41%). One of these patients with high‐grade HPV‐related VIN developed SCC after 4 years, the epithelium being surrounded by HPV‐positive undifferentiated VIN. The other two patients with progressive disease developed SCC with differentiated VIN not related to HPV in the adjacent skin (table 11).
This is the first study describing the histology and HPV status of a large series of patients with both LS and VIN without coexistent SCC. In this study, VIN associated with LS was always of undifferentiated type. HPV DNA was detected in eight of 26 (31%) patients, seven of whom had high‐grade VIN.
It is commonly thought that there is a relationship between differentiated VIN and LS.17,18 There are, however, only few data available on this subject, and almost all these concern the adjacent skin of vulvar SCC (table 22).4,5,7,9,19,20,22,28,29 In 153 of 467 (33%) patients whose skin lesions in the surrounding tissue of vulvar SCC were analysed, the existence of both VIN and LS has been observed (range 5–58%). Predominantly, this VIN was of differentiated type (74%). Sporadically, the coexistence of VIN and LS has been described in vulvar skin without invasive carcinoma (table 33).19,20,21,22 Since these studies were on either LS or VIN, mostly of undifferentiated type, selection bias might have affected the results.
To date, the presence of HPV DNA in VIN related to LS has been described only once.19 Haefner et al reported on three patients with HPV‐negative differentiated VIN, all in the adjacent skin of vulvar SCC, and on three patients with HPV‐positive undifferentiated VIN 3, in one of whom VIN was adjacent to SCC and in the other two it was not. The detection of HPV DNA in 41% of our high‐grade VIN lesions supports the diagnosis of undifferentiated VIN, although, compared with the prevalence of HPV DNA in VIN not related to LS (78–92%), this percentage is rather low.19,30,31,32 This lack of similarity between undifferentiated VIN with and without associated LS might point to a different pathogenesis, at least in half of the cases.
It is known that there are some difficulties in the histological diagnosis of VIN. Firstly, it can be very difficult to distinguish VIN 1 from atypical inflammatory reactive changes or (damaged) normal skin, resulting in a high interobserver variability.33,34,35 In our study, we excluded 10 of 46 patients, in whom low‐grade VIN diagnosed in the original histology report was not confirmed at revision following the definition given by the ISSVD.15 The high prevalence of VIN 1 lesions that tested negative for HPV after revision (89%) gives the impression that we still overdiagnosed VIN. However, MIB 1 staining, which has been suggested to be useful in accurate grading of VIN,35 supported the diagnosis of VIN 1 in all cases. Today, this VIN 1–3 classification is the subject of discussion, and a new classification in which the term VIN 1 will no longer be used has been proposed by the ISSVD. Similarly, Medeiros et al recently recommended a distinction between low‐grade VIN (VIN 1) and high‐grade VIN (VIN 2 or 3 and differentiated VIN) to identify those lesions at risk for vulvar carcinoma.36 In our study, such a new classification would affect the number of patients considerably, causing a significant increase in the number of HPV‐related high‐grade VIN associated with LS (43%). Secondly, the presence of differentiated VIN has to be considered when there seems to be only a slight degree of squamous atypia or it is limited to the lower epidermis. Because of its highly differentiated features and absence of widespread architectural disarray, the diagnosis can be easily missed. We specifically looked into the subtle but characteristic features of differentiated VIN, as described in Materials and methods. However, we did not observe any of these features, except during follow‐up, surrounding early invasive SCC. Furthermore, it may be difficult to agree upon whether early invasive growth has already occurred in differentiated VIN.25 Conversely, differentiated VIN is easily diagnosed when there is an invasive SCC. The fact that none of our revised VIN 1 cases progressed to invasion strengthens our claim of not having missed any differentiated VIN.
As we studied only those cases with both LS and VIN from the start, a selection bias might have been caused by the design of our study. Revising all LS cases over several years could result in a higher prevalence of LS with coexisting VIN, and therefore possibly in a higher number of differentiated VIN. Revising all VIN cases, on the other hand, might result in a higher number of coexistent diseases as well, with VIN most likely to be of undifferentiated type.
In conclusion, it seems from an overview of the literature that the assumed relationship between differentiated VIN and LS can be supported in the case of coexistent vulvar SCC,4,7,19,20,29 but has been described in only five patients without SCC.20,22 In this study, we found that VIN related to LS without SCC was always of the undifferentiated type. These data suggest that VIN in the background of LS without SCC is a different type of VIN than that was thought previously, with a different aetiology and perhaps a different prognosis. In that case, we might question ourselves whether or not we tend to overtreat patients if we radically excise VIN in the background of LS. In our opinion, treatment of VIN associated with LS should be individualised, based on type and localisation of the lesion, in order to achieve a better conservation of the anatomy and function of the vulva.
HPV - human papillomavirus
ISSVD - International Society for the Study of Vulvovaginal Diseases
LS - lichen sclerosus
SCC - squamous cell carcinoma
VIN - vulvar intraepithelial neoplasia
Competing interests: None.