p16INK4A expression in squamous cell carcinomas of the vagina and the vulva in Tunisian women.

BACKGROUND
The role of p16INK4A expression in uterine cervix cancer is well established. In the remaining female lower genital tract cancers, the importance of p16INK4A up-regulation is less clear. In our study, we analyzed the role of p16INK4A expression and HPV infection in carcinomas of the vulva and the vagina in Tunisian women.


MATERIALS AND METHODS
We conducted a retrospective study of 30 carcinomas including 15 vulvar squamous cell carcinomas (SCCs) and 15 vaginal SCCs. Immunohistochemistry was used to determine p16INK4A expression. HPV detection and typing was by in situ hybridization.


RESULTS
p16INK4A expression was detected in 86.7% of vaginal SCCs with a strong and diffuse immunostaining in 60% of cases, and also in 73.3% of vulvar SCCs with focal immunoreactivity in 53.3% The association between p16INK4A expression and HPV infection was significant in vaginal SCCs (p=0.001) but not vulvar SCCs (p>0.05).


CONCLUSIONS
p16INK4A expression could be used as a useful marker for HPV positivity in vaginal SCCs similar to that described in uterine cervix cancers. However, our data support the presence of 2 different mechanisms for p16INK4A expression in HPV-related and HPV-unrelated vulvar carcinomas.


Introduction
Cancers of the female genital tract represent the third most common malignant neoplasms in women after cancers of the breast and of the digestive tract in Tunisia (Missaoui et al., 2010a). The most common types of female genital tract cancers are cervical, ovarian and endometrial carcinoma. There are other less common tumors including tumors of vagina, vulva and fallopian tube. Neoplasms of the vulva and vagina account for less than 5% of all female genital tract cancers (Franco, 1996;Benedet et al., 2000;Wells et al., 2003). Squamous cell carcinomas (SCCs) represent the most common malignant tumors, accounting for more than 85% in both localizations (Del Pino et al., 2013). Vaginal and vulvar SCCs may have many of the same risk-factors as uterine cervix SCCs, including the association with persistent human papillomavirus (HPV) infection (Merino, 1991;Carter et al., 2001;Daling et al., 2002;zur Hausen, 2002;Hellman et al., 2004;Ferreira et al., 2008;Wu et al., 2008;Siriaunkgul et al., 2014). HPV infection has been detected in 40% of vaginal cancers and HPV16 was the most HPV type detected (Daling et al., 2002;Hampl et al., 2006;Parkin et al., 2006;Srodon et al., 2006). Over the last two decades, two different etiopathogenic pathways for the development of vulvar SCCs and intraepithelial RESEARCH ARTICLE
Previously, we supported the role of p16 INK4A overexpression as a useful additional marker for the interpretation of problematic uterine cervix lesions reducing the variability during evaluation of suspicious biopsies (Missaoui et al., 2010b). More anteriorly, we considered that p16 INK4A is a putative molecular biomarker that consistently discriminates uterine cervix adenocarcinomas from benign lesions and from endometrioid adenocarcinomas of the uterine corpus (Missaoui et al., 2006).
The role of p16 INK4A expression in the remaining female lower genital tract cancers is less studied because of the rarity of these tumors. Recently, Alonso et al. considered that p16 INK4A immunostaining can be easily implemented in routine pathology and should be considered as valuable prognostic biomarkers of vaginal cancers (Alonso et al., 2012). In 1998, Chan et al. reported an increasing p16 INK4A expression with the vulvar lesion grade and they suggested that p16 INK4A alterations could be significant events in progression of vulvar disease (Chan et al., 1998). More recently, Knopp et al. analyzed the p16 INK4A expression in larger series of vulvar SCCs and they reported a significant correlation between the high p16 INK4A expression and a better prognosis in the multivariate analysis and less risk of developing lymph node metastasis (Knopp et al., 2004).
In the current study, we analyzed the role of the p16 INK4A expression and the HPV infection in vaginal and vulvar SCCs cancers among Tunisian patients.

Tissue samples
We carried out a retrospective study of 30 female lower genital tract cancers retrieved from the surgical pathology files of the Department of Pathology, Farhet Hached University Hospital, Sousse, Tunisia. These cases were selected randomly. All slides were reviewed by two pathologists (Dr. Sihem Hmissa and Dr. Moncef Mokni). Ethical approval for use of all specimens was obtained from the research ethics committee of the Farhet Hached University Hospital.
The cases studied were distributed into the following groups, according to the World Health Organization (WHO) Classification of Tumors of the Breast andFemale Genital Organs, 2003 (Wells et al., 2003): 15 vulvar squamous cell carcinomas (SCCs) and 15 vaginal SCCs. All tissues had been routinely fixed in 4% buffered formalin and paraffin-embedded.

Immunohistochemistry for p 16INK4A protein expression
The immunostaining procedure was carried out as we already described (Missaoui et al., 2006;Missaoui et al., 2010b;. Briefly, one or two paraffin blocks containing representative portions of the cancers were selected for each case and 4 μm-thick serial sections were obtained. Sections were incubated for 30 min with primary monoclonal antibodies against anti-p16 INK4A protein (Dako Cytomation, K5334, clone E6H4, dilution 1:50). The remaining part of the procedure was performed as we previously published (Missaoui et al., 2006;Missaoui et al., 2010b;. One invasive uterine cervix carcinoma with known diffuse and strong immunoreactivity with p16 INK4A antibody was used as a positive control. Negative controls, using monoclonal mouse immunoglobulin G (IgG2a) antibody at a comparable concentration, were included.

In situ hybridization for HPV infection
The in situ hybridization technique was carried out as already described (Nabi et al., 2006;Hachana et al., 2010). Briefly, one or two paraffin blocks containing representative portions of the cancers were selected for each case and 3μm-thick serial sections were obtained by microtome. A wide spectrum biotinylated probe for common HPV types was used according to the manufacturer's suggested protocol (Dako GenPoint K0620, Dako, Carpinteria, California, USA). The wide spectrum probe (Y1404) targets the genomic DNA of HPV types 6, 11, 16, 18, 30, 31, 33, 35, 45, 51, and 52. Further HPV typing was carried out on cases found to be positive using the wide spectrum probe using specific probe for HPV16/18 (Y1412) according to manufacturer's protocol. Two uterine cervix cancer cases were used as positive control cases that were positive in previous reactions.

Statistical analysis
Data analyses were carried out using the Epi-Info 2002 software as previously described (Missaoui et al., 2006). The association between p16 INK4A expression and HPV infection was analyzed by Chi-square statistics. Probability values of 0.05 or less were considered statistically significant.

Results
Immunostaining for p16 INK4A protein In our study, no immunoreactivity for p16 INK4A protein was detected in normal areas adjacent to cancerous lesions. p16 INK4A expression was detected in 86.7% of vaginal SCCs cases. No immunoreactivity for p16 INK4A protein was detected in 2 vaginal SCCs (13.3%). There was strong and diffuse nuclear and cytoplasmic p16 INK4A immunostaining of the neoplastic cells in 60% of vaginal SCCs with score 12 (Figure 1A-B). In the 20% of vaginal SCCs cases, p16 INK4A expression was focal not exceeding a score of 4 (Table 1). p16 INK4A expression was observed in 73.3% of vulvar SCCs, whereas no p16 INK4A expression was detected in the remaining 4 vulvar SCCs cases (Table 1). The p16 INK4A immunoreactivity was focal and scattered, not exceeding a score of 3 in 53.3% of all vulvar SCCs ( Figure 1C). However, strong and diffuse immunoreactivity for p16 INK4A (score 12 and 15) was observed in both the nucleus and cytoplasm in only 2 cases ( Figure 1D).  (Figure 2A). Only 3 cancer cases were HPVnegative. The HPV16/18 infection was observed in 38.5% of vaginal SCCs HPV-positive (5 cases) ( Figure 2B).
Among vulvar SCCs, HPV infection was detected in 40% of studied cases ( Figure 2C). The remaining cases were HPV-negative by in situ hybridization (9 cases). HPV16/18 infection was present in only one case (16.7%).

Discussion
During the last years, p16 INK4A expression has been considered as a surrogate marker for HPV-positive uterine cervix cancers (Benevolo et al., 2006;Vinyuvat et al., 2008;Kurshumliu et al., 2009;Missaoui et al., 2010b;Cheah et al., 2012;Genoves et al., 2014). The role of p16 INK4A expression in the remaining female lower genital tract cancers is less studied. In this study, we extensively analyzed the immunohistochemical distribution of p16 INK4A protein expression in SCCs of the vagina and the vulva and the association with HPV infection.    (Hellman et al., 2014). Strong p16 INK4A expression was significantly correlated with low histopathological grade, HPV positivity, and long-term survival. Their findings indicated that p16I NK4A and Ki-67 expression might be useful in tumor grading and p16 INK4A expression could be used as a useful marker for HPV positivity in vaginal carcinoma (Hellman et al., 2014). Currently, p16 INK4A expression was observed in 73.3% of vulvar SCCs. However, no significant association was observed between p16 INK4A and HPV infection in these tumors. The p16 INK4A expression was detected in HPVpositive (33.3%) and HPV-negative (40%) vulvar SCCs. Our results clearly support previous studies suggesting the presence of 2 different mechanisms for p16 INK4A expression in HPV-related and HPV-unrelated vulvar cancers (Riethdorf et al., 2004;der Avoort et al., 2006;Hoevenaars et al., 2008;de Sanjosé et al., 2013). Riethdorf et al. analyzed for p16 INK4A expression and HPV infection by RNA/RNA in situ hybridization in a large series of vulvar lesions (Riethdorf et al., 2004). These researchers considered that, as in the uterine cervix, intense diffuse p16 INK4A expression supports an HPV-related neoplastic process in vulvar neoplasia, irrespective of the level of differentiation. However, the up-regulation of p16 INK4A at the epithelial-stromal interface in HPV-negative keratinizing SCCs is consistent with an HPV-independent response to alterations associated with invasion. In this regard, the disparate patterns of p16 INK4A expression underscore 2 different mechanisms for p16 INK4A expression in HPV-related and HPV-unrelated vulvar carcinomas.
Interestingly, van der Avoort et al. (2006) provided further evidence that vulvar SCCs is a multifactorial disease that develops from two different pathways. First, an HPV-dependent pathway with a remarkable resemblance to uterine cervix carcinomas and second, an HPV-independent pathway in which differentiated vulvar intraepithelial neoplasia III lesions that are high-risk HPVnegative, may be precursors.
Recently, a worldwide study of HPV infection in  (Rufforny et al., 2005). The researchers investigated the expression of p16 INK4A protein and the detection of HPV16 by real-time PCR in 49 vulvar lesions including benign/reactive lesions, condyloma acuminatum, vulvar intraepithelial neoplasia, and invasive SCCs. Although, the up-regulation of INK4A gene occurs in vulvar carcinogenesis, p16 INK4A expression is not a sensitive marker for differentiation of benign vulvar squamous epithelium from condyloma acuminatum or VIN 1 lesions. They considered that p16 INK4A expression may aid in the diagnosis of HPV-related lesions and as such may be of value as a surrogate marker in the diagnosis of vulvar premalignant and malignant lesions (Rufforny et al., 2005).
A histologic study of 92 vulvar SCCs was conducted to evaluate the usefulness of p16 INK4A immunohistochemistry in the classification of vulvar SCCs (Santos et al., 2006). Diffuse p16 INK4A expression was observed in all HPV-positive vulvar SCCs and in only 2.3% of HPV-negative cases. The sensitivity and specificity of p16 INK4A immunostaining to detect HPV-associated carcinomas (100% and 98.7%, respectively) were better than those of histologic criteria (93.8% and 35.5%). Moreover, no differences in age, stage, or development of recurrence were observed between HPV-positive and negative tumors. These finding supported the significant overlapping of the morphologic criteria to discriminate HPV-positive and negative vulvar SCCs. In this regard, Santos et al. considered that the p16 INK4A immunostaining is a reliable marker for HPV-positive tumors, which improves the results of histologic classification of vulvar SSC (Santos et al., 2006).
Moreover, the prognostic significance of p16 INK4A expression in invasive vulvar SCCs was investigated by Tringler et al. (2007). The expression was localized to the cytoplasm and the nuclei of 43% of tumor cells. p16 INK4A -positive patients showed a significantly longer disease-free and overall survival by univariate analysis. p16 INK4A staining may be of prognostic significance in invasive vulvar SCCs.
In summary, our results clearly support the role of p16 INK4A expression and HPV infection in the carcinogenesis of vagina similar to that described in the uterine cervix cancer. In this regard, p16 INK4A expression should be regarded as a surrogate biomarker of vaginal SCCs and HPV infection. However, in the vulva, the status of p16 INK4A expression observed in our study suggests the