Purpose
To evaluate the role of high-risk human papillomavirus (HR-HPV) infection in periocular sebaceous carcinoma (SC) using multiple methods of detection, and to determine whether p16 overexpression is present and can be used as a surrogate marker for HR-HPV.
Design
Retrospective observational case series with laboratory investigations.
Methods
Unstained paraffin sections of 35 cases of periocular SC were analyzed with immunohistochemistry for p16 and subjected to polymerase chain reaction (PCR) for HR-HPV. A subset of 18 lesions that were p16-positive was further studied with a novel method of mRNA in situ hybridization (ISH) for the detection of transcriptionally active HR-HPV, an advanced technique with an enhanced sensitivity and specificity.
Results
The clinical findings were in keeping with those of comparable earlier studies. Strong immunohistochemical p16 positivity (meeting the criterion of >70% nuclear and cytoplasmic staining) was present in 29 of 35 cases of periocular SC (82.9%). The selected 18 p16-positive cases tested were negative for HR-HPV using mRNA ISH. PCR yielded unequivocal results with adequate DNA isolated in 24 cases, 23 of which were negative for HR-HPV. One case was positive for HPV type 16, which was found to be a false positive as collaterally determined by mRNA ISH negativity.
Conclusion
No evidence was found for HR-HPV as an etiologic agent in the development of periocular SC using multiple modalities to maximize sensitivity and specificity and reduce the limitations of any single test. p16 overexpression is common in periocular SC but unrelated to HR-HPV status. Although p16 may be used as a surrogate marker for HR-HPV status in other tissue sites, this interpretation of p16 positivity is not applicable to periocular SC.
The reported incidence of high-risk human papillomavirus infection (HR-HPV) in periocular sebaceous carcinoma (SC) has ranged from 0 to 57%. Most studies either were based on small numbers of patients and/or used only 1 method of viral detection (eg, in situ hybridization [ISH] alone or polymerase chain reaction [PCR] alone). In parallel, p16 overexpression has consistently been reported in a high percentage of SC, regardless of HPV status (if evaluated). In certain other anatomic sites such as the oropharynx, p16 positivity in nonkeratinizing squamous cell carcinomas is used as a surrogate marker for HR-HPV, obviating additional viral studies.
In this study, 35 cases of periocular SC were analyzed for the presence of HR-HPV using the immunohistochemical probe p16 with PCR. Eighteen of these cases were also evaluated with a novel and highly sensitive method of mRNA ISH that has not been previously employed in ophthalmic investigations. The specific research objectives in the current investigation were 2-fold: (1) to establish whether or not a credible linkage exists between HR-HPV infection and the pathogenesis of SC by limiting confounding factors that could be responsible for underdetection or contamination; and (2) to investigate whether the reported high rate of p16 expression in periocular SC correlates with HPV positivity.
Methods
Case Selection and Slide Review
This retrospective, observational case series was conducted under the auspices of the Massachusetts Eye and Ear Infirmary (MEEI) and the Massachusetts General Hospital (MGH) Institutional Review Board, in compliance with the rules and regulations of the Health Insurance Portability and Accountability Act and all applicable federal and state laws, and in adherence to the tenets of the Declaration of Helsinki. A search of the MEEI/MGH pathology information system was performed. Sequential surgical samples of 20 periocular SC diagnosed over a 16-year period (1998–2015) at MEEI/MGH were selected. An additional 9 cases were collected and contributed by the Emory University Hospital and 12 cases by the New York Eye and Ear Infirmary (NYEEI).
The formalin-fixed, paraffin-embedded (FFPE) tissue blocks and glass slides of these 41 cases were retrieved from the pathology archives of the 3 aforementioned academic hospitals. Hematoxylin-eosin–stained sections were reviewed to confirm the histopathologic diagnosis and assess the histomorphologic features of each tumor. Immunohistochemistry for cytoplasmic lipid was performed using adipophilin to confirm the diagnosis. At this stage, 6 cases were excluded owing to the absence of characteristic histopathologic findings for sebaceous carcinoma or negative staining for adipophilin in a vesicular pattern. The remaining 35 cases formed the basis for investigative study. Unstained, 5-μm-thick FFPE sections from the 35 surgical samples were subsequently evaluated for the presence of HR-HPV by p16 immunohistochemistry (Leica Biosystems, Wetzlar, Germany) and PCR (Qiagen, Hilden, Germany). A subset of 18 cases was evaluated using mRNA-ISH.
p16 Immunohistochemistry
Immunohistochemical expression of the cyclin-dependent kinase inhibitor p16 was evaluated in all cases. In brief, deparaffinized FFPE sections of all cases were subjected to antigen retrieval using the Leica Bond protocol (Leica Biosystems, Wetzlar, Germany) with proprietary Retrieval ER2 (ethylene diamine tetraacetic acid solution, pH 9.0) for 20 minutes. A mouse monoclonal antibody against p16 (E6H4 clone, CINtec; Ventana Medical Systems, Tucson, Arizona, USA) was used with a 1:4 dilution, detected by the Polymer Refine Kit (Leica Biosystems, Wetzlar, Germany) on a Leica Bond Autostainer. For positive immunohistochemical controls, a tonsil squamous cell carcinoma with positive p16 expression was used. The threshold for p16 positivity was met in cases where ≥70% of tumor cells demonstrated strong diffuse nuclear and cytoplasmic staining; other staining patterns were considered negative.
High-Risk Human Papillomavirus Status by Polymerase Chain Reaction
Regions of interest were macro-dissected from 5-μm sections of FFPE tissue to enrich for tumor DNA content. For polymerase chain reaction–restriction fragment length polymorphism (PCR-RFLP), DNA was isolated using QIAamp Mini kit (Qiagen, Hilden, Germany) following manufacturer’s instruction. HPV genotyping was done by PCR-RFLP method. Briefly, a PCR reaction was first performed using HPV consensus primers designed to amplify a conserved 332-to470-bp fragment of the HPV L1 gene and PCR product analyzed on a 5% polyacrylamide gel stained with ethidium bromide. If a visible product was present, the PCR product was digested with 3 different restriction enzymes, Pst I, Rsa I, and Hae III. The digested products were visualized on a 5% polyacrylamide gel stained with ethidium bromide and the patterns of the digested products compared to a database of known patterns to determine the genotype. A control PCR reaction with primers to a 500-bp fragment of the human beta globin was also performed to assess the quality of the DNA. Alternatively, HPV genotyping was performed using the HPV Linear Array assay (Roche Diagnostics, Indianapolis, Indiana, USA). For this assay, tumor was similarly enriched using macro-dissection from 5-μm tissue sections and DNA extraction was performed with the Promega Maxwell FFPE extraction kit (Promega, Fitchburg, Wisconsin, USA) according to manufacturer’s protocol. HPV Linear Array was performed according to manufacturer’s protocol. The sample internal control (beta-globin gene target) was used to confirm successful PCR amplification.
Automated RNA In Situ Hybridization Assays: Pooled High-Risk Human Papillomavirus and Cocktail Human Papillomavirus 16/18 Probes
An automated RNA ISH assay was performed using pooled probes for the HPV16 and HPV18 E6 and E7 mRNA. The assay was performed using the View-RNA eZL Detection Kit (Affymetrix) and BDZ 6.0 software (Leica Biosystems). Briefly, 5 mm-thick sections of formalin-fixed tissue were baked for 1 hour at 60 C and placed on the Bond RX for processing. The Bond RX user-selectable settings were as follows: ViewRNA eZ-l Detection 1-plex (Red) protocol; ViewRNA Dewax1; View RNA HIER 10 minutes, ER1 (95); ViewRNA Enzyme 2 (10); ViewRNA Probe Hybridization. With these settings, the RNA unmasking conditions consisted of a 10-minute incubation at 95 C in Bond Epitope Retrieval Solution 1 (Leica Biosystems), followed by 20-minute incubation with proteinase K from the Bond Enzyme Pretreatment Kit at 1:1000 dilution (Leica Biosystems). HPV type 16 (cat# DVF1-17255) and HPV 18 (cat# DVF1-17256) were pooled and diluted 1:40. ViewRNA Probe Diluent (Affymetrix). Post run, slides were rinsed with water, air dried for 30 minutes at room temperature, and mounted using Dako Ultramount (Dako, Carpinteria, California, USA), and were visualized using a standard bright-field microscope. Presence of punctate dot-like red-colored hybridization signals in the tumor cells was defined as HR-HPV-positive tumor.
Results
The studied cohort of periocular SC consisted of 35 cases comprising 12 men and 23 women with a mean age of 78.9 years (range, 46–94 years). Tumors more commonly presented in the upper eyelid ( Figure 1 , Left and Right). In 13 biopsy specimens, the tumor was confined to the epithelium ( Figure 2 , Top left) of the eyelid skin or conjunctiva; in the remaining 22 biopsies there was invasive disease ( Figure 2 , Top right). Ten cases contained only invasive disease in the areas sampled. Intraepithelial disease was found in the form of either carcinoma in situ or pagetoid spread of single cells. Some tumors were composed of basaloid cells containing inconspicuous vacuoles and others were composed of sebocyte-like cells (well-differentiated cases) with well-formed vacuoles and/or a frothy appearance of the cytoplasm. Comedonecrosis was a common feature of invasive tumors ( Figure 2 , Middle left). Significant nuclear pleomorphism was observed, including the presence of bizarre nuclei and many mitotic figures per high-power field. Vesicular cytoplasmic adipophilin positivity was uniformly present ( Figure 2 , Middle right). Immunohistochemistry for p16 was definitively positive in 82.9% (29/35) of tumors ( Figure 2 , Bottom left and right). The threshold for p16 positivity was >70% of cells with both nuclear and cytoplasmic staining. This level was used because it best correlates with HPV status in nonophthalmic head and neck squamous tumors. Five of 6 of the p16-negative cases contained invasive carcinoma; 1 case was entirely intraepithelial in its limited sampling. Another 3 tumors were focally or weakly positive for p16 and were categorized as p16-negative. PCR for HR-HPV was performed on all 35 tumors and was negative in 24 of 25 cases with adequate tumor DNA and unequivocal results; in 10 others, results were inconclusive or insufficient DNA was isolated (7 of these cases were greater than a decade old). One instance of invasive SC was positive for HPV type 16 by PCR. A subset of 18 (p16-positive) cases was studied with mRNA ISH, including the PCR-positive case, and all 18 were negative ( Table 1 ).
| Study | Number of Cases | Method(s) of HPV Detection | Other Markers Investigated | Results |
|---|---|---|---|---|
| Hayashi et al, 1994 (Japan) | 21 eyelid SC | 1. DNA ISH | p53 (IHC) | 13/21 HPV positive (61.9%) 12/21 HR-HPV positive (57.1%) |
| Gonzalez-Fernandez et al, 1998 (Virginia, USA) | 7 eyelid SC | 1. RNA ISH 2. PCR | p53 | 0/7 HPV positive (0%) |
| Cho et al, 2000 (South Korea) | 18 eyelid SC | 1. DNA ISH | ER, PR, c-erbB2 | 0/18 HPV positive (0%) |
| Liau et al, 2014 (Taiwan) | 24 periocular SC | 1. PCR | p16, CDKN2A , mismatch repair proteins | 1/24 HPV positive (4.2%) |
| Kwon et al, 2015 (South Korea) | 14 eyelid SC | 1. PCR | p16, KRAS, HER2 | 0/12 HPV positive (0%) |
| Current study (Massachusetts, USA) | 35 periocular SC | 1. PCR (24 cases) 2. mRNA ISH (18 cases) | p16 (35 cases) | 0/24 HPV positive (0%) |
Results
The studied cohort of periocular SC consisted of 35 cases comprising 12 men and 23 women with a mean age of 78.9 years (range, 46–94 years). Tumors more commonly presented in the upper eyelid ( Figure 1 , Left and Right). In 13 biopsy specimens, the tumor was confined to the epithelium ( Figure 2 , Top left) of the eyelid skin or conjunctiva; in the remaining 22 biopsies there was invasive disease ( Figure 2 , Top right). Ten cases contained only invasive disease in the areas sampled. Intraepithelial disease was found in the form of either carcinoma in situ or pagetoid spread of single cells. Some tumors were composed of basaloid cells containing inconspicuous vacuoles and others were composed of sebocyte-like cells (well-differentiated cases) with well-formed vacuoles and/or a frothy appearance of the cytoplasm. Comedonecrosis was a common feature of invasive tumors ( Figure 2 , Middle left). Significant nuclear pleomorphism was observed, including the presence of bizarre nuclei and many mitotic figures per high-power field. Vesicular cytoplasmic adipophilin positivity was uniformly present ( Figure 2 , Middle right). Immunohistochemistry for p16 was definitively positive in 82.9% (29/35) of tumors ( Figure 2 , Bottom left and right). The threshold for p16 positivity was >70% of cells with both nuclear and cytoplasmic staining. This level was used because it best correlates with HPV status in nonophthalmic head and neck squamous tumors. Five of 6 of the p16-negative cases contained invasive carcinoma; 1 case was entirely intraepithelial in its limited sampling. Another 3 tumors were focally or weakly positive for p16 and were categorized as p16-negative. PCR for HR-HPV was performed on all 35 tumors and was negative in 24 of 25 cases with adequate tumor DNA and unequivocal results; in 10 others, results were inconclusive or insufficient DNA was isolated (7 of these cases were greater than a decade old). One instance of invasive SC was positive for HPV type 16 by PCR. A subset of 18 (p16-positive) cases was studied with mRNA ISH, including the PCR-positive case, and all 18 were negative ( Table 1 ).
