Prevalence of human papillomavirus and Epstein–Barr virus in salivary gland diseases

Abstract

Objective

The roles of human papillomavirus (HPV) and Epstein–Barr virus (EBV) in head and neck neoplasms have been well reported, but little is known about their relationship with salivary gland tumours. This study investigated the presence of HPV and EBV in salivary gland diseases.

Methods

The presence of HPV 16/18 and EBV was analysed in archival pathological specimens collected from patients who had undergone surgery for salivary gland diseases. HPV 16/18 DNA was detected using nested polymerase chain reaction (PCR) and further confirmed with immunohistochemistry. EBV DNA was detected using real-time PCR.

Results

A total of 61 pathological specimens were examined: 39.5% (15/38) of pleomorphic adenomas, 33.3% (3/9) of Warthin’s tumours, 33.3% (one of 3) of mucoepidermoid carcinomas, and 25.0% (one of 4) of benign lymphoepithelial lesions were positive for high-risk HPV 16/18. Only two Warthin’s tumours were positive for EBV.

Conclusion

The infectious nature of salivary gland neoplasms was revealed by the high prevalence of HPV infection, and the specific presence of EBV in Warthin’s tumours, suggesting a potential role for HPV and EBV in salivary gland diseases.

Introduction

The implication of viruses as aetiological factors in head and neck benign and malignant tumours is a focus of attention for biomedical research. Studies in head and neck neoplasms have specifically investigated Epstein–Barr virus (EBV) and human papillomavirus (HPV).^1–5 EBV is an enveloped DNA herpesvirus that is transmitted by saliva and is shed even in apparently healthy subjects. ¹ In head and neck cancers, the role of EBV in tumourigenesis has been demonstrated for undifferentiated carcinoma of the nasopharynx,^2–4 and has been shown to be quantifiable. ⁵ Not only was EBV found in malignant neoplasms, it was also detected in benign lesions such as lymphoepithelial cysts. ⁶ HPV has been shown to be involved in the aetiology of epithelial cell cancers, such as cervical squamous cell carcinoma. ⁷ In the head and neck region, HPV has recently been suggested to have a relationship with oropharyngeal cancers.^8–10 Benign neoplasms, such as laryngeal papilloma, were reported to have a relatively higher incidence of malignant transformation in Asian patients compared with other ethnicities. ¹¹ Patients without demonstrable HPV DNA showed significantly higher cumulative risk of malignant transformation than HPV-positive patients, probably due to an activated antiviral cellular immune response in the HPV-positive group. ¹¹

Neoplasms of salivary origin, mostly benign, may be encountered in various locations in the head and neck. Comprising the majority of all salivary gland neoplasms, pleomorphic adenoma or mixed tumour, is a salivary gland neoplasm that has the potential for recurrence and malignant transformation. ¹² The second most frequently encountered benign salivary gland tumour, papillary cystadenoma lymphomatosum (or Warthin’s tumour) also has a propensity for multifocal or bilateral occurrence. ¹³ The role of viruses in the tumourigenesis of salivary gland tumours remains unclear. The purpose of this current investigation was to determine the prevalence of HPV and EBV infection in salivary gland neoplasms. The involvement of these two tumour-related viruses was analysed using polymerase chain reaction (PCR), direct sequencing, and real-time PCR in a spectrum of archival salivary gland specimens, including both benign and malignant tumours. The presence of high-risk HPV viral proteins in salivary gland tumours was also confirmed with immunohistochemistry.

Patients and methods

Patients

This retrospective study identified appropriate patients through a hospital electronic chart data system using coding for discharge diagnosis. The study considered consecutive patients who underwent surgery for salivary gland diseases at the Department of Otolaryngology and the Department of Surgery, Tungs’ Taichung Metro Harbour Hospital, Taichung, Taiwan between 1 January 1998 and 31 December 2007. Consecutive retrieval of tissue samples based on the site of the tumour originating from the salivary glands, completeness of data recorded in the hospital charts and surgical date were performed, starting with patients who were more recently processed for conventional histological procedures. Pathological tissue samples deemed to possess preserved integrity and completeness were retrieved from the pathology archives of Tungs’ Taichung Metro Harbour Hospital. Exclusion criteria were patients who had not received surgery or who had an insufficient pathological sample for the study. The study was approved by the Institutional Review Board of Tungs’ Taichung Metro Harbour Hospital (IRB 97011, IRB 99008 and IRB 100027). Patient consent was not required for analysis of archival material.

Preparation of specimens and DNA isolation

Formalin-fixed paraffin wax-embedded specimens were sectioned at a thickness of 5 µm. Blades were disinfected with alcohol for each sample and the microtome was adjusted to ensure that a new area of cut surface was being used for each block. All sections were deparaffinized in xylene, rehydrated through a graded series of alcohol, and washed in 10 mM phosphate-buffered saline (PBS; pH 7.2). ¹⁴ Genomic DNA was prepared from a tissue section and isolated by conventional phenol–chloroform extraction, followed by ethanol precipitation, then dissolved in 30 µl sterile distilled water. Amplifiable quality of DNA was confirmed by PCR targeting the housekeeping gene human glyceraldehyde-3-phosphate dehydrogenase. All tests included a negative control amplification containing double-distilled, autoclaved, filtered water and PCR reagents.

Nested PCR for HPV detection

For HPV detection in tissue sections, genomic DNA prepared as described above was used for nested PCR reactions. Amplification of HPV DNA was first performed with type-consensus primers, MY09 and MY11. ¹⁴ This was followed by a second round of amplification with type-specific primers flanking the L1 region to identify the HPV 16 and HPV 18 subtypes, as reported previously. ¹⁴ Briefly, 5 μl of each sample was amplified with MY09/11 primers (10 pmol each) and 2 U of AmpliTaq® DNA polymerase (Applied Biosystems, Foster City, CA, USA) using Molecular BioProducts EasyStart™ PCR tubes (1 × PCR buffer II, 2 mM MgCl₂, 200 μM [each] of dATP, dCTP, dGTP, and dTTP; Thermo Fisher Scientific, Rockford, IL, USA). Additional MgCl₂ was added to each reaction to a final concentration of 4 mM. Amplifications were performed in a MyCycler™ thermal cycler (Bio-Rad, Hercules, CA, USA) using the following settings: preliminary denaturation at 95℃ for 2 min, followed by 40 cycles of denaturation at 95℃ for 1 min, annealing at 55℃ for 1 min, and elongation at 72℃ for 1 min. This was followed by a final elongation step at 72℃ for 10 min and storage at 4℃. After loading 10 μl of the final PCR product onto a 2% agarose gel stained with ethidium bromide, the electrophoresed product was visualized under ultraviolet illumination and photographed simultaneously. In each PCR reaction, appropriate negative and positive controls were included. A part of the human glyceraldehyde-3-phosphate dehydrogenase gene in all samples was amplified to exclude false-negative results, whereas DNA preparations from SiHa cells (containing HPV 16) and HeLa cells (containing HPV 18) were used as positive controls.

Direct sequencing for HPV subtype confirmation

The confirmation of HPV 16 and HPV 18 subtypes was determined by direct sequencing of PCR products amplified from the DNA of specimens using a BigDye® Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems) and an Applied Biosystems ABI PRISM® 3100/3100 Avant Genetic Analyzer (Applied Biosystems). Target sequences were amplified using the same type-specific primers flanking the L1 region for PCR. The resulting sequences were analysed in the Basic Local Alignment Search Tool data bank for comparison with known HPV sequences. ¹⁵

Immunohistochemistry for confirmation of HPV protein

For immunohistochemistry studies, tissue sections cut to a thickness of 3 µm were deparaffinized in xylene, sequentially rehydrated in alcohol, and washed three times in 10 mM PBS (pH 7.2). Sections were heated three times for 5 min each on a low setting in a 1800-watt microwave oven (CR-1800 microwave oven; Sampo, Taoyuan, Taiwan) in 10 mM citrate buffer (pH 6.0) to facilitate antigen retrieval. Sections were then incubated with polyclonal anti-HPV 16 or 18 E6 antibody (Santa Cruz Biotechnology, Santa Cruz, CA, USA; dilution 1 : 200) for 90 min at 25℃. The conventional streptavidin peroxidase method (LSAB™ Kit; Dako, Glostrup, Denmark) was performed to develop the signals and the cells were counterstained with haematoxylin. Negative controls, obtained by leaving out the primary antibody, were performed for every set of experiments to ensure the quality of the assay. The signals were evaluated independently by two observers (P-L.C. & T-Y.T.) and scored by the presence of positive nuclei: score 0, no positive staining; score +, positive staining. Positive control slides for cervical cancer tumour tissues with HPV 16/18 were used as a positive control for HPV 16/18 E6. The detailed protocols were described in previous reports.^14,16

Real-time PCR for EBV detection

For EBV detection in tissue sections, DNA was employed for real-time PCR amplification of the BamHI-W region of the EBV genome, using forward and reverse primers as previously described. ¹⁷ Following a similar protocol and reaction setup procedures, the same dual fluorescence-labelled oligomer was used as previously described. ¹⁷ Fluorogenic PCR reactions were added in a reaction volume of 50 µl using components supplied in a TaqMan® PCR Core Reagent Kit (Perkin Elmer, Foster City, CA, USA). Fluorescent probes were custom-synthesized by Perkin Elmer. PCR primers were synthesized by Life Technologies (Paisley, UK). Each reaction contained 5 µl of 10 × buffer A; 300 nM of each of the amplification primers; 25 nM(for the EBV probes) or 100 nM(for the β-globin probe) of the corresponding fluorescent probe; 4 mM MgCl₂; 200 µM each of dATP, dCTP, and dGTP; 400 µM dUTP; 1.25 U of AmpliTaq Gold® (Applied Biosystems); and 0.5 U of AmpErase® uracil-N-glycosylase (Applied Biosystems). Extracted DNA (5 µl) was used for the amplifications. DNA amplifications were carried out in a 96-well reaction plate format in a StepOne™ Real-Time PCR System (Applied Biosystems). Each sample was analysed in duplicate. Multiple negative controls using water blanks were included in every analysis. All DNA samples were subjected to internal control for the RNase P gene using TaqMan® RNase P Control Reagents Kit (Applied Biosystems).

Statistical analyses

All statistical analyses were performed using SPSS®, version 17.0 (SPSS Inc., Chicago, IL, USA) for Windows®. χ²-test was used to compare categorical variables between the salivary gland tumour group and the group of other salivary gland diseases, between the Warthin’s and non-Warthin’s tumour groups, and between the HPV 16 and HPV 18 groups. Significance was established at the 5% (P ≤ 0.05) level for all statistical analyses.

Results

There were 108 consecutive patients who underwent surgeries for salivary gland diseases; of these, 61 pathological tissue samples were considered to be suitable for analysis because they possessed preserved integrity and completeness. Of the 61 patients included in the study, 32 were female and 29 were male. Their mean ± SD age was 44.7 ± 16.0 years (range 19–86 years) (Table 1). Fifty-seven (93.4%) neoplasms involved the major salivary glands (the parotid and submandibular glands).

Table 1.

Demographic and clinical characteristics of patients (n = 61) who underwent surgery for salivary gland diseases and were included in a study to investigate the role of Epstein–Barr virus and human papillomavirus in the development of benign and malignant neoplasms.

Characteristic	Patients with salivary gland diseases
	Salivary gland tumours n = 53	Other salivary gland diseases n = 8
Age, years	44.4 ± 6.9	41.8 ± 10.1
Male	24 (45.3)	5 (62.5)
Female	29 (54.7)	3 (37.5)
Primary site
Parotid gland	41 (77.4)	6 (75.0)
Submandibular gland	8 (15.1)	2 (25.0)
Nasal	2 (3.8)	0 (0.0)
Hard palate	2 (3.8)	0 (0.0)

Data presented as mean ± SD or n of patients (%).

Tissue samples obtained from 50 benign salivary gland tumours, three malignant salivary gland tumours, and eight samples from other salivary gland diseases were analysed for the presence of HPV and EBV (Table 2). Of those with positive findings for HPV 16/18 DNA tested by nested PCR, all were subjected to direct sequencing for subtype confirmation. All HPV 16/18 DNA-positive cases were further confirmed with immunohistochemistry for the presence of HPV 16/18 viral protein (Figure 1). Thirty-eight pleomorphic adenomas from 53 salivary gland tumours were tested for the presence of both HPV 16 and 18 DNA (Table 2). HPV 18 was detected in significantly more pleomorphic adenomas than HPV 16 (12 of 38 [31.6%] versus 3 of 38 [7.9%] specimens, respectively; P = 0.003). HPV 18 was detected in one female patient who underwent total parotidectomy for a second recurrence of pleomorphic adenoma. When the salivary gland tumours and the samples from other salivary gland diseases were compared, a higher proportion of the tumours were HPV 16/18 positive, although no statistical significance was demonstrated. Three out of nine patients (33.3%) with Warthin’s tumours were positive for HPV 16/18. In the eight other salivary gland diseases, only one benign lymphoepithelial lesion was positive for HPV 18. Only two of nine (22.2%) Warthin’s tumours and none of the other lesion types had detectable EBV DNA (P = 0.003 for Warthin’s tumours versus non-Warthin’s tumours). There were two cases of nasal septum pleomorphic adenomas, but neither tumour was positive for EBV or HPV 16/18 DNA.

Figure 1.

Representative photomicrographs showing positive immunostaining for human papillomavirus (HPV) 16 (a) and HPV 18 (b) in paraffin wax-embedded sections of pleomorphic adenoma. Sections were counter-stained with haematoxylin. The colour version of this figure is available at: http://imr.sagepub.com. Scale bar 50 µm.

Table 2.

The presence of Epstein–Barr virus (EBV) and human papillomavirus (HPV) detected by nested polymerase chain reaction (PCR) and real-time PCR in samples from patients (n = 61) who underwent surgery for salivary gland diseases.

Diagnosis	n (%)	HPV		EBV
		16	18
Salivary gland tumours	53 (86.9)	4 (7.5)	15 (28.3)	2 (3.8)
Pleomorphic adenoma	38 (62.3)	3 (7.9)	12 (31.6)	0 (0.0)
Warthin’s tumour	9 (14.8)	1 (11.1)	2 (22.2)	2 (22.2)
Basal cell adenoma	2 (3.3)	0 (0.0)	0 (0.0)	0 (0.0)
Mucoepidermoid carcinoma	3 (4.9)	0 (0.0)	1 (33.3)	0 (0.0)
Myoepithelial tumour	1 (1.6)	0 (0.0)	0 (0.0)	0 (0.0)
Other salivary gland diseases	8 (13.1)	0 (0.0)	1 (12.5)	0 (0.0)
Benign lymphoepithelial lesion	4 (6.6)	0 (0.0)	1 (25.0)	0 (0.0)
Sialadenosis	2 (3.3)	0 (0.0)	0 (0.0)	0 (0.0)
Reactive hyperplasia	1 (1.6)	0 (0.0)	0 (0.0)	0 (0.0)
Cellular schwannoma	1 (1.6)	0 (0.0)	0 (0.0)	0 (0.0)
Total	61	4	16	2

Data presented as n (%) of patients.

Discussion

Advances in molecular biology and its application in medical research have enabled the detection of viral DNA in human pathological specimens archived for >10 years. A role for HPV and EBV in the pathogenesis of head and neck neoplasm has gained interest. The role of HPV in head and neck cancers, especially in oropharyngeal cancers, has been reported in several studies.^9,18 Similarly, the role of HPV in salivary gland lesions has been examined. In a group of nine parotid lesions studied by Vageli et al., ¹⁹ seven were HPV-positive whereas six were infected with high-risk HPV 16/18. Another study showed that eight neoplastic salivary gland lesions from 34 specimens (four Warthin’s tumours, two pleomorphic adenomas, one myoepithelioma and one lymphoma) were associated with HPV 16/18. ²⁰ In contrast, a study of 19 pleomorphic adenomas and 19 malignant salivary gland tumours reported that HPV was not detected in any of the tumour samples by PCR. ²¹ Two further studies reported that either none of 55 or two of 38 salivary gland malignancies were HPV 16/18 positive by PCR or in situ hybridization.^22,23 This present study demonstrated that 35.8% (19/53) of salivary gland tumours were positive for HPV 16/18. Nested PCR was used to improve the detection of HPV DNA in the present study,^14,16 and the presence of viral protein was confirmed using immunohistochemistry studies. Specifically, the detection of HPV 16/18 was 39.5% (15/38) in the pleomorphic adenomas tested. To our knowledge, this is the largest number of pleomorphic adenomas associated with high-risk HPV infection ever reported in the current literature regarding salivary glands. Furthermore, when the salivary gland tumours and the samples from other salivary gland diseases were compared, a higher proportion of tumours was HPV 16/18 positive, although no statistical significance was demonstrated, implying HPV involvement in tumourigenesis.

The current results showed a significantly higher prevalence of HPV 18 than HPV 16 in pleomorphic adenomas (P = 0.003). This observation differed from those made in other head and neck squamous cell carcinomas, where HPV 16 was most commonly found.^18,24 The reported prevalence of HPV 16 in oropharyngeal cancers was 38.2% and 58.6%.^18,24 In addition to nested PCR, this current study also performed direct DNA sequencing to confirm the HPV 16/18 subtypes. The current results showed 19 out of 53 (35.8%) salivary gland tumours were positive for HPV 16/18: 4 (7.5%) with HPV 16; and 15 (28.3%) with HPV 18. The implication of HPV 18 subtype predominance in HPV-infected salivary gland tumours will need to be determined in future investigations.

After pleomorphic adenoma, Warthin’s tumour is the second most common benign salivary gland tumour reported, ¹² which was also observed in this current study. EBV was only found in Warthin’s tumours (22.2%), consistent with results from previous studies.^21,25 Although the presence of EBV has been demonstrated previously in benign lymphoepithelial cysts of the parotid gland, ²⁶ the four cases of benign lymphothelial lesions in this current study were not positive for EBV. The presence of EBV in nasal pleomorphic adenoma has been described in the literature, mostly occurring in the nasal septum.^21,27 In this current investigation, there were two cases of nasal septum pleomorphic adenomas, but neither tumour was positive for EBV or HPV 16/18 DNA. Nasal septum pleomorphic adenoma is a rare entity that was first reported in 1932 and until now it has been presented mostly as case reports.^21,27–30 Therefore, further studies (including more cases) are warranted to determine whether EBV is involved in the aetiology of pleomorphic adenomas of the nasal septum.

In this current series, one of three patients (33.3%) with mucoepidermoid carcinoma was positive for HPV 18. Atula et al. ²¹ reported that none of 19 malignant salivary gland tumours, including three mucoepidermoid carcinomas, was positive for HPV. However, another study demonstrated that HPV 16/18 was detected in two mucoepidermoid tumours from 38 salivary gland malignancies by in situ hybridization. ²³ As described earlier, the current study employed a more sensitive method of HPV detection compared with previous research; ²³ this method has been applied to the detection of HPV in other diseases under similar laboratory conditions.^14,16 The presence of high-risk HPV in malignant salivary gland tumours has been reported in parotid lesions, although the number of samples was small, with a total of only nine parotid lesions being included. ¹⁹

In conclusion, this current study provides insight into the prevalence of high-risk HPV and EBV infection in salivary gland diseases in the general adult population. This study demonstrated that 39.5% of pleomorphic adenomas, 33.3% of Warthin’s tumours, 33.3% of mucoepidermoid carcinomas and 25.0% of benign lymphoepithelial lesions were positive for high-risk HPV 16/18. A higher proportion of the HPV 18 subtype was found in the salivary gland lesions compared with HPV 16. Of the 61 patients, only two cases of Warthin’s tumour tested positive for EBV. To our knowledge, this study has the largest series of HPV or EBV detection in salivary gland diseases reported in the literature. The high detection rates suggest the possible involvement of these two viruses in salivary gland diseases.