BSCC of the uterine cervix is a rare subtype as defined by the most recent WHO classification. BSCC is considered an aggressive squamous cell carcinoma variant, which has received little attention in the literature. [6] It is difficult to make a distinct diagnosis of BSCC from biopsy specimens because there are various spectrums of high-grade basaloid tumors in the cervix, particularly ACC, which may be similar. Some papers have reported that BSCC in the anus and esophagus displayed an adenoid cystic pattern, which resembled ACC of the head and neck. [3, 7] The hyalinized stroma around the tumor, eosinophilic basement membrane, and cystic changes can be confused with solid, high-grade ACC. The correct diagnosis from BSCC to ACC is important because of its worse prognosis.
ACC is also a rare tumor arising most commonly in the salivary and lacrimal glands and rarely other organs, such as the skin, breast, lung, uterine cervix, and vulva. [8] ACC in the uterine cervix has been reported to be associated with high-risk HPV infection in some cases.[9] However, data between p16 expression and HPV infection in high-grade ACC is limited.
In case 2, it was especially difficult to distinguish BSCC from ACC with a solid variant. It showed a cribriform pattern of basaloid cell proliferation with variable amounts of myxoid material. Additional sections were obtained from the hysterectomy specimen for further evaluation. The periphery of the tumor showed a basaloid cell nest consisting of a palisading arrangement with distinct squamous differentiation (Fig. 1E). The immunoprofile of the tumor showed diffuse reactive staining for p63 (Fig. 1G) and p16 (Fig. 1M), but was immunonegative for S100 protein (Fig. 1J) unlike ACC. BSCCs displayed diffuse immunoreactivity for p63 whereas ACC showed a compartmentalized pattern within tumor nests, in which p63-positive cells were found internally as well as at the periphery.
Squamous cell carcinoma of the uterine cervix, including basaloid squamous cell carcinoma, is associated with HPV infection. Diffuse strong positivity of p16 immunohistochemical staining is a well-known surrogate marker for carcinoma associated with HPV infection. The p16 gene (INK4a/CDKN2A) is a tumor suppressor gene that encodes a cyclin-dependent kinase inhibitor involved in cell cycle regulation and which inhibits Rb phosphorylation leading to cell cycle arrest and a negative feedback loop to control cell proliferation. During high-risk HPV infection, DNA integrates into the genome of the host squamous cells, the viral oncoprotein E7 binds to the retinoblastoma protein, and induces its functional inactivation. The overexpression of p16 appears due to the loss of negative feedback.[10]
Diffuse strong p16 overexpression has been described in ACC of the head and neck; however, the association of HPV infection with ACC is controversial. Boland et al. reported that only 11% of high-grade ACCs of the head and neck were combined with high-risk HPV infection among high grade ACC showing immunopositivity for p16. These results suggested that the presence of HPV infection in high-grade ACC was not paralleled by the strong expression of p16. [11] Although ACC of the uterine cervix may be accompanied by other HPV-related lesions[12], p16 immunopositivity does not always reflect HPV infection status in ACC. p16 overexpression can result from different mechanisms independent of HPV, such as CDKN2A gene mutations, deletions, and promoter hypermethylation.[13] All three cases of ‘basaloid carcinoma of the uterine cervix’ underwent immunohistochemistry for p16 staining and molecular analysis by HPV real-time PCR in this study. All three cases showed diffuse strong expression of p16. The presence of high-risk HPV was revealed in 2 BSCC cases; however, HPV infection was not detected in the ACC case.
There are distinctive ultrastructural features between BSCC and high-grade ACC by electron microscopy. The polygonal cells of BSCC and ACC had a thick basement membrane and reduplicated external lamina. Also, there was no distinctive differentiation in nuclear morphology. In BSCCs from cases 1 and 2, tumors showed relatively undifferentiated cells consisting of cystic lumen without their luminal microvilli; however, ACC showed occasional ductal differentiation with luminal microvilli.
In conclusion, diffuse strong p16 expression in cervical cancer is generally associated with HPV infection. However, it is not a reliable surrogate marker of HPV in relation to ACC. Therefore, the use of other special methods is important to differentiate basaloid SCC from high-grade ACC. In cervical cancer, additional HPV PCR or mRNA in situ hybridization should be used to differentiate HPV-associated squamous cell carcinoma.
HPV analysis and immunostainings for p63 and S100 should be useful adjuncts to diagnose the ‘basaloid’ carcinoma of the uterine cervix. It may also be helpful to observe true ductal formation with microvilli by electron microscopy.