Diffuse p16 Expression is Against the Diagnosis of DEK::AFF2 Carcinoma

To the Editor:

We read the recent article entitled “Sinonasal DEK‐rearranged Papillary Non‐keratinizing Squamous Cell Carcinoma: Expanding the Emerging Entity” published in the Indian Journal of Otolaryngology and Head & Neck Surgery with some doubt about the conclusion [1]. In this article, Laforga et al. reported a papillary non-keratinizing sinonasal squamous cell carcinoma (SNSCC) presenting as a polypoid mass in the left middle turbinate of a 53-year-old female. The carcinoma was basaloid, arranged in variably thickened trabeculae with pseudoglandular structures, and without keratinization or papillae. It was diffusely positive with immunohistochemical stains for pancytokeratin, CK5, p63, and p16 (with a strong block staining pattern), while negative for INSM1, NUT, and NSE. Staining for INI-1 and SMARCA4 immunohistochemistry (IHC) showed both retained expression. In situ hybridization for human papillomavirus (HPV) type 16/18 (unspecified DNA or RNA probe) was negative making the authors conclude the result of p16 immunohistochemistry (IHC) as non-contributory. They reported the DEK break-apart fluorescence in situ hybridization (FISH) analysis as positive for the rearrangement.

Our first concern is about the interpretation of the FISH result. In their Figure 5, the DEK break-apart FISH analysis showed only overlapping red and green signals or fused yellow signals, which suggested an unequivocally negative result for DEK rearrangement and precluded the diagnosis of DEK::AFF2 carcinoma. A positive DEK break-apart FISH should demonstrate split red and green signals in > 20% of nuclei in 50 non-overlapping tumor cells, as seen in our previous study [2]. The second concern is about the strong block staining pattern of p16 IHC in this tumor, which fulfilled the criteria of the College of American Pathologist guidelines (moderate to strong nuclear and cytoplasmic expression in ≥ 70% of tumor cells) as a positive surrogate marker of high-risk HPV in head and neck carcinomas [3]. In the previous studies including ours, a total of 20 DEK::AFF2 carcinomas were tested for p16 IHC and only 50% (10/20) of them showed a mosaic pattern of expression in < 70% of tumor cells [2, 4]. None of these cases had a block p16 expression pattern. Although the p16 IHC result was deemed non-contributory by Laforga et al. due to a negative HPV 16/18 in situ hybridization, it has been recognized that other high-risk HPV types, aside from 16 and 18, are associated with carcinomas arising in the sinonasal tract. Therefore, we implore the authors to consider HPV-related multiphenotypic sinonasal carcinoma (HMSC) as the main differential diagnosis. This comment is based on the alleged history of “epithelial-myoepithelial carcinoma” in the patient’s prior biopsy and the morphologic findings of a basaloid tumor forming anastomosing trabeculae and adenoid cystic-like pattern, as shown in their Figure 3B. HMSC may display morphologic overlap with DEK::AFF2 carcinoma and is only rarely associated with HPV types 16 and 18 [5]. To confirm this possibility, it is prudent to recommend HPV genotyping to include types 33 and 35 along with additional IHC, such as actin or calponin, to demonstrate myoepithelial differentiation.

In summary, it is highly unlikely for the reported case to be a DEK::AFF2 carcinoma given the negative DEK break-apart FISH result and the diffuse block p16 expression pattern. Consequently, we urge the authors to perform further work-up for other pertinent differential diagnoses, including HMSC.

Declarations

Conflict of interest

The authors have no conflict of interest.