4 Discussion

Poor reproducible diagnostic criteria and ambiguous immunostaining patterns, along with morphologic discordance still affects the accurate diagnosis and classification of vulvar precancer lesions, especially dVIN and dVIN related vSCCa. To address this issue, we studied a series of 60 first-time diagnosed dVIN-related vSCCa from our archived pathologic records. The majority of our cases (80%) are morphologically moderately to poorly differentiated. dVIN-related vSCCa usually occurred in postmenopausal women with peak age of 70–80 years. Forty-six cases (77%) were classified as FIGO stage 1B and above according to the AJCC 8th Edition.

One of the most important morphologic features of dVIN is vulvar basal layer atypia and the high degree of differentiation. By histology, the basal layer of vulvar skin show “cytologic atypia” with vesicular nuclei with thickened, irregular nuclear membranes, coarse chromatin/hyperchromasia, and prominent nucleoli. Other histologic features of dVIN include the increased mitosis, particularly atypical mitosis in the basal cell layer, elongation and anastomoses of rete ridges, dyskeratosis, and prominent acantholysis or spongiosis. Abnormal keratinization and reverse maturation along with basal cell proliferation may also be identified. These features can be extremely subtle, which results in low interobserver agreement even among experienced gynecologic pathologists.^4,19,20 There are some occasions where it can be challenging to distinguish between dVIN and uVIN, especially when dVIN presents with a more basaloid appearance extending beyond the basal layer, architectural disorganization, and undifferentiated keratinocytes.^4,21 In these instances, ancillary immunohistochemical studies can be helpful, such as the hrHPV surrogate marker, p16. In our study, all of the p16 positive cases had negative HPV RNA in-situ hybridization results, regardless of whether p16 was positive in dVIN or vSCCa.

TP53 gene mutation has been reported in up to 80% of dVIN, however its role in dVIN pathogenesis as well as malignant transformation is still being understood. Recently, six p53 immunohistochemical staining patterns have been described, in which the authors have shown its strong correlation with TP53 mutation status.²² In this study, the authors identified six major p53 IHC patterns, two wild-type patterns: (1) scattered, (2) mid-epithelial expression (with basal sparing), and four mutant patterns: (3) basal overexpression, (4) para-basal/diffuse overexpression, (5) absent expression, and (6) cytoplasmic expression. These patterns were consistent with TP53 mutation status in 95% vSCCa and 93% in-situ lesions. They concluded that those exhibited scattered p53 staining and those with a p53 weak basal overexpression could be easily confused. Rakislova et al²³ compared patterns of p53 positivity and identified a new six-pattern framework between invasive vSCCa and an adjacent skin lesion. They found that approximately 74% invasive vSCCa showed abnormal p53 staining and 47% vSCCa with associated skin lesions showed an abnormal p53 stain, with an identical staining pattern between the vSCCa and the adjacent skin lesion in 80% of the cases.

In the present study, we identified four major staining patterns with respect to p53 and p16 expression in dVIN related vSCCa: 1) pattern of “mutant p53 and negative p16” in dVIN and invasive carcinoma (44/60); 2) “wild-type p53 and negative p16” (3/60); 3) “mutant p53 and block positive p16” (2/60); 4) “null p53 and negative p16” (5/60). Additionally, we identified atypical staining patterns in six cases in our cohort: null p53 in dVIN but overexpression in invasive carcinoma (2/60); null p53 in invasive carcinoma with two showing basal and para-basal p53 in dVIN (3/60); wild-type p53 pattern in dVIN (1/60). Only one case demonstrated p53 wild-type pattern in dVIN but overexpression in invasive carcinoma.

TP53 mutations are widely accepted as supportive of a diagnosis of dVIN. Somatic missense, deletion, or truncating mutations of the TP53 tumor suppressor gene are commonly found in HPV-negative vulvar carcinogenesis, although a subset of them may develop through TP53-independent routes. Petitjean et al have reported that strong and diffuse p53 expression is a surrogate for TP53 missense mutation, whereas TP53 deletion or truncating mutations are associated with a null phenotype,²⁴ although the null phenotype of p53 has not been well studied in the vulva yet.

Noojj et al demonstrated that the TP53 mutations were much more common in HPV-independent precancer lesions than those related with HPV infection.²⁵ They further identified a novel, third genetically distinct group of tumors that were HPV-independent but p53 wild-type (so-called p53 ​wt VIN) but were enriched for somatic mutations in the NOTCH1 and HRAS genes. However, they found no difference in rates of local recurrence or 5-year survival between HPV negative/p53 ​wt and HPV negative/p53 mutant vSCCa in their study.^26–28

It may be difficult to interpret p53 expression patterns in some cases. Leigl et al have reported the p53 overexpression in lichen sclerosus, therefore p53 is not specific for a diagnosis of dVIN.²⁸ They also proposed that wild-type p53 protein may accumulate following cell stress or DNA damage, a mechanism postulated to occur in lichen sclerosus where ischemic stress results from the poor oxygenation, further resulting in vasculitis and inflammation.²⁹ In addition, wild-type and mutant p53 staining pattern may overlap, further complicating interpretation.

This is an observational clinical study with an inclusion of a large number of dVIN-related vSCCa cases from a single academic women's hospital. Both in-situ and invasive carcinoma components along with an immunohistochemical stain panel including p16, p53, and Ki67 were clinicopathologically reviewed. All strongly positive p16 IHC cases were confirmed hrHPV status by RNA in-situ hybridization. Studies have shown that p53 IHC pattern has an excellent correlation with TP53 sequencing, albeit it was not performed in current study.²⁴ This study corroborates previous findings that the p53 expression patterns are concordant in dVIN and the subsequent dVIN related vSCCa.²⁴ However, there is a need for additional studies addressing the mutational profiles of dVIN and dVIN-related vSCCa with discordant or discrepant p53 expression patterns, especially in cases with unusual morphologic features. In this regard, we recommend the clinical correlation and morphologic re-evaluation. An immunohistochemical panel including p53, Ki67, and p16 should be utilized for a diagnosis of dVIN-related vSCCa.