Low Epstein-Barr virus count in sinonasal inverted papilloma

Abstract

Background: Sinonasal inverted papilloma (SIP) is a benign tumour originating from the sinonasal mucosa showing an extensive growth pattern, a high risk of recurrence and a 5–10% risk to malignify. Epstein-Barr virus (EBV) is an oncogenic herpesvirus which infects most individuals via the saliva eliciting a latent infection. Previous studies have been reporting variable data on EBV in SIP, and there is no present appreciation regarding the association between these.

Aims/objectives: The aims were to investigate the presence and count of EBV in SIP and map the viral distribution in the epithelium versus the connective tissue.

Material and method: Fifty-three SIP patients were identified in the Pathology Department register at the University Hospital of Umeå. The biopsies were analysed with Epstein-Barr Encoded Region (EBER) in situ hybridization. EBER-positive cells were counted in the epithelium and connective tissue.

Results: We found EBER-stained cells in 30% of the cases, where 19% of these had an abundance of stained cells, and the rest showed a low count.

Conclusions/significance: These findings demonstrate a low EBV count in SIP. EBV is less likely to be a causative agent in the formation of SIP, or its malignant transformation.

Chinese abstract

背景：鼻腔内翻性乳头状瘤（SIP）是一种起源于鼻腔粘膜的良性肿瘤, 具有广泛的生长模式、高复发风险和5-10%的恶性转化风险。Epstain-Barr病毒（EBV）是一种致癌的疱疹病毒, 通过唾液传染给大多数人, 引起潜在感染。先前的研究报告了SIP中EBV的变量数据。目前对这些数据之间的关系还没有认识。

目的：研究SIP中EBV的存在和计数, 并绘制病毒在上皮和结缔组织中的分布图。

材料与方法：在Umeå大学医院病理科登记处确认53例SIP患者。采用EBER原位杂交技术对活检组织进行分析。计数上皮和结缔组织中的EBER阳性细胞。

结果：30%的病例中发现EBER染色细胞, 其中19%的病例染色细胞丰富, 其余病例计数低。

结论/意义：这些发现表明SIP的EBV计数较低。EBV不太可能是SIP形成或恶性转化的诱因。

Keywords:

• Epstein-Barr virus
• sinonasal inverted papilloma
• EBER-ISH

Introduction

Sinonasal inverted papilloma (SIP) is a benign tumour originating from the sinonasal mucosa. It shows an extensive growth pattern, a high risk of recurrence and a 5–10% risk to become malignant [1]. Transformation into squamous cell carcinoma (SCC) can occur synchronously with SIP or develop metachronously at a site with previous benign SIP. Clinically, this means a need for thorough removal of SIP-affected mucosa with a technologically-enhanced, endoscopy-led, trans-nasal approach [2].

The incidence of SIP has been estimated to 0.2–1.5/100,000 person-years [3]. SIP comprises 0.5–4% of primary nasal tumours without any right or left side predilection. The most common sites of origin are the following, in order of descending prevalence: the lateral nasal wall, the ethmoid cells, and maxillary sinus. According to Anari et al., the male to female ratio is between 3:1 and 5:1, with a peak age of presentation in the fifth and sixth decades [4]. The SIP diagnosis is generally made at a late stage, 1–4 years after first onset of sinonasal symptoms. The most common symptoms are nasal obstruction, anterior and/or posterior rhinorrhoea, headache with facial pressure feeling, or a combination of multiple symptoms [5].

Clinical examination with endoscopy of the nasal cavities can reveal a pink-greyish, firm lobulated tumour, with a characteristic “raspberry” aspect. Biopsy for histological examination of SIP shows epithelium inverting into the stroma with a distinct and intact basement membrane that separates and defines the epithelial component from the underlying connective tissue stroma. The epithelium provides a clear-cut demarcation to respiratory mucosa, and lacks mucus-secreting cells and eosinophils [4].

Though the exact aetiology of SIP is not known, some findings indicate that viral agents including human papilloma virus (HPV), cell cycle and angiogenic factors, chronic inflammation, environmental and occupational exposure to organic solvents and welding fumes, have been associated with this condition [3].

Epstein-Barr virus (EBV) is a gamma herpesvirus infecting children and adolescents via the saliva. EBV is associated with cellular hyperproliferative disorders such as infectious mononucleosis, B cell- and solid cell tumors [6]. The oral EBV antigens cross the epithelial barrier in the oropharynx, and are presented to naïve B cells in the underlying tissue. The primary EBV infection gives rise to a persistent latent infection due to its interaction with memory B cells [7]. In the germinal centre, EBV biology mirrors B cell biology, and EBV-infected memory B cells downregulate virus gene expression to avoid immune recognition.

Earlier reports show divergent results for whether or not EBV has an etiological role in the development of SIP [8,9]. This divergence in findings might be explained by differences in inclusion criteria, sample sizes and methodology.

An EBV infection in patients is usually detected by polymerase chain reaction (PCR), and it is also a rapid and accurate method for identifying EBV DNA-copies in tissue biopsies. However, PCR technique, neither detects how many individual cells are EBV-infected, nor does it show the location of these cells. In this study, we used Epstein Barr Encoding Region In Situ Hybridization (EBER-ISH), which reveals staining in the nuclei of the EBV-infected cells. EBER-ISH is a good marker of EBV infection, since large numbers of copies of EBERs are present in latently infected cells and are easily identified [10]. Although the function of EBERs is not clear, Greifenegger et al. reported that in latently EBV infected cells the viral genes EBER-1 and EBER-2 are transcribed by polymerase III, and that in lytically infected cells in vivo the EBER genes are not detected. [11].

The purpose of this study was to investigate the presence and count of the potentially oncogenic EBV in SIP. A specific aim was to map the viral distribution in the epithelium versus the connective tissue.

Material and method

Material

In this cross-sectional study, we included 88 patients diagnosed with inverted papilloma (SIP) in the nasal cavity and paranasal sinuses between 1984 and 2014. The samples analysed were obtained with permission from the Biobank North (County Council of Västerbotten) approval number 472-13-008, with latest update 2019-04-25. The patients gave verbal consent for use of the biopsy tissue for banking in the pathology register and research before undergoing surgery/biopsy. The study was approved by the Regional Ethical Review Board of Umeå University (approval number 2017-543-32 M, addendum to 2015-323-32 M, 2012-379-32 M, and [2010-277-31M], approval date 2018-01-29.

Patient tissue biopsies were collected from the pathology register, at the department of Clinical Pathology, University Hospital of Umeå, by ICD-O-3 codes for inverted papilloma of the nose and paranasal sinuses (80531/80531) and exophytic nasal papilloma/inverted urothelial cell papilloma (81210). A new histopathological analysis was performed to guarantee that exophytic nasal papilloma/inverted urothelial cell papilloma, sometimes coded as inverted papilloma, were not included in the study.

This study is an extension of a previous publication from our research group, on the oncogenic human papilloma virus and p16, using the same SIP material, by Holm et al. [12]. From 54 biopsies left, one biopsy was excluded due to lack of material, and 53 biopsies were eligible and analysed with EBER-ISH.

Patients characteristics

The gender distribution in patient samples analysed with EBER-ISH (n = 53) was 70% men and 30% women. The age range was 18–88 years, with a mean of 59 years.

Method

The EBER-ISH is a commercially available in situ hybridization kit for EBER transcripts (Epstein-Barr Encoding Region).

Two tissue sections were cut out from each SIP biopsy, and one section was incubated with a RNA-positive control probe (800-2846, Ventana) to make sure RNA was available in the sample. The other section was tested for EBER-1 and EBER-2 by the Inform EBER Probe (800-2842, Ventana Medical Systems, Roche Diagnostics GmbH, Mannheim Germany). For visualization, NBT-BCIP Detection System was used (800-092, Roche Diagnostics GmbH, Mannheim, Germany). The staining was performed in a Bench Mark Ultra (Ventana Medical Systems, Inc, Tucson, AZ, USA) according to manufacturer’s guidelines.

EBER-positive cells were counted by two of the authors (AS, KN), both in epithelium and connective tissue of the SIP. Consensus was reached after comparison of the results.

In this material we classified the finding of EBER stained cells into either presence or absence of cells of a tissue slide (Figure 1). All cells in the tissue slides were studied.

Figure 1. Example of EBV presence. Dark blue cells representing EBER positivity in the epithelium of SIP in a tissue slide.

Statistics

Descriptive statistics (IBM SPSS USA, version 25) are presented. Cross tabulation is used to describe the relationship between numbers of EBER-positive cells in the epithelium and connective tissue. Correlation tests were not feasible due to too few positive cases.

Results

EBV presence and distribution

All EBER analysis control slides were positive, confirming method quality.

Out of 53 patients with SIP, 4 patients (7.5%) had one or more EBER-positive cell in the epithelium, 13 patients (24.5%) had positive cells in the connective tissue, and one had positive cells in both epithelium and connective tissue (1.9%) (Table 1). The outcome measures were binary, either presence or absence of positive cells.

Table 1. EBV found in the connective tissue.

	Negative	Positive	Total
EBV found in the epithelium
Negative	37 (70%)	12 (23%)	49
Positive	3 (6%)	1 (1.9%)	4
Total	40	13	53

Of those with EBER-positive cells in the connective tissue, three patients showed high expression. Two of these had mild inflammation, with 20% and 50% EBER positivity, respectively. The third patient had >50% EBER positive cells along with a high grade of inflammation. The remaining patients had 1–15 EBER positive cells. Two out of the three patients with high EBER cell expression were younger (29 and 44 years) than the group mean age. Patients with low EBER expression (1–15 positive cells in the connective tissue), had a mean age of 64 years, median age of 66 years, and 30% were women and 70% men.

Of patients who expressed positive cells in the epithelium, all had between 1 and 10 positive cells each (per biopsy slide). Mean age was 53.5 years, median age was 46 years and gender distribution was 25% women and 75% men.

Samples, including both epithelium (4/53) and connective tissue (12/53), were EBER positive in 30% of the SIP cohort. EBV expression was low, 1–15 cells in 81% (13/16) of cases, and high (20%, 50%, >50%) in 19.3% (3/16) of cases.

Discussion

The purpose of this study was to investigate the presence of EBV in SIP. The aims were to map the EBV distribution in the epithelium versus the connective tissue. We found EBV in 30% of the cases, with a distribution of 7.5% in the epithelium and 24.5% in the connective tissue.

SIP is most likely characterized by a multi-mediator and -step aetiology with a strong tendency to recur and grow invasively despite its benign histology [4]. This strongly implicates an infectious aetiology. HPV has been suggested as a potential aetiological agent for sinonasal papillomas and their associated malignancies [13]. The role of EBV as a recognized oncovirus in the development of SIP is still unclear, and also if concurrency of HPV and EBV can enhance promotional events in a multistep oncogenesis. Our research group has previously published an HPV and p16 analysis on the same biopsy material [12], and we found no co-expression between HPV and EBV.

In a Swedish population-based study, it was shown that the incidence of SIP has increased from 0.01 to 0.33/100,000 person-years from 1960 to 2010 [14]. Also, the risk of squamous-cell carcinomas in the SIP cohort is significantly higher than in the general population, with a 300-fold increase among male patients with SIP.

Sinonasal inverted papillomas were EBER positive

Fifty-three SIP patients were assessed for eligibility and tissue samples retrieved from Clinical Pathology at the University Hospital of Umeå between 1984 and 2014. There are different interpretations regarding expression of EBER, though no definition or scoring system for low numbers of EBER expressing cells is available. Regarding B-cell lymphomas, EBER-positivity is defined as >20% or >50% [15]. There is also no validated rank-system described for low count of EBER positive cells. We decided to rank the findings of EBER expressing cells into either absence or presence of cells due to the large number of slides without positive cells and to facilitate descriptive statistics. We could present the percentage of samples with positively stained cells in the epithelium and connective tissue, which was in total 30%. Expression of EBER was low (1–15 cells) in 81% of cases. In three cases (19%) expression was high in the connective tissue, one with 20% and two with at least 50% EBER-positivity.

It is not widely agreed if presence of EBV in epithelial cells can be normal, but according to the Germinal Centre Model [7] infection of epithelial cells is pathological. EBV-associated epithelial cancers include nasopharyngeal carcinoma and gastric carcinoma. Doescher et al. found that 45% (20/44) of sinonasal squamous cell carcinomas were positive for EBV with a significantly more aggressive disease outcome [16].

Choice of method for EBV detection

To assess the number of EBV-infected cells, and map the distribution of EBER positive cells, we chose in situ hybridization (ISH) versus polymerase chain reaction (PCR). Epstein-Barr Encoded Regions 1 and 2 (EBERs) are small RNAs that are regarded as reliable markers for ISH to detect EBV infection in clinical tissue samples such as nasopharyngeal carcinoma, lymphomas and gastric carcinoma. EBERs are thought to aid in the activation of oncogenic signalling pathways [17].

In one study Macdonald et al. [8] could detect EBV in 65% of SIP specimens by the use of PCR. Later, other authors, like Sham et al. [9], concluded that EBV was absent in SIP using both PCR and EBER-ISH. They also acknowledged that the role of EBV in the pathogenesis of SIP remains controversial due to reports showing divergent prevalence of EBV in SIP ranging from 0% to 68%.

Other possible causes of sinonasal inverted papilloma

For our cohort there were no data collected on industrial exposure to solvents or fumes, smoking or other exogenous irritants that can be related to SIP. A retrospective case-control study from Italy showed that the risk of inverted papilloma was significantly associated with ever exposure to organic solvents and welding fumes, with a dose-response relationship observed in organic solvents [18].

Smoking is one of the most important risk factors for neoplasms in the head- and neck mucosa. But it is not considered a causative agent in SIP formation according to analytical data, but may contribute to recurrence and malignant transformation of SIP [3].

Preventive vaccination from oncogenes

Prophylactic vaccine against EBV has been a topic of research under the last decades, mainly focusing on the EBV major envelope protein gp350. No licensed EBV vaccine is currently available. It has been suggested that prevention of infectious mononucleosis could decrease the disease burden of Hodgkin lymphoma and multiple sclerosis [19] as well as several other malignancies of hematopoietic or epithelial origin.

It is important to study the HPV status in specified upper airway locus since specific HPV genotypes are prevented by vaccination. Vaccination can be used as treatment in specific cases in order to reduce surgical intervals, and care-givers need to promote this in their treatment planning. There has been a clear-cut decline in recurrent respiratory papillomatosis (RRP) induced by HPV 6 and 11 in Australia, following the implementation of a national vaccine program [20].

Methodological considerations/limitations

We chose in situ hybridization (ISH) to assess the number of EBV infected cells and map the distribution of EBER positive cells, versus polymerase chain reaction (PCR) that detects EBV DNA copies.

Due to low count of low and high EBV expression in epithelium and connective tissue respectively, we were not able to statistically compare the locations and quantities. Small sample size affects the power negatively and increases the risk of type II errors. The result is instead presented using cross tabulation for comparison between numbers of EBER-positive cells in the epithelium and connective tissue with numbers, percentage and medians. The retrospective design with little data collection beyond the pathology specimens is also a limitation in that we could not identify other associated exposure factors.

Conclusion

The findings demonstrate Epstein-Barr virus in sinonasal inverted papilloma using EBER-ISH assessing cellular presence, distribution and count.

We found EBER-stained cells in 30% of cases distributed with a quarter in the connective tissue and only 7.5% in the epithelium, which is the location of malignification. It might be of pathological value but cannot be used to infer a cause- and effect relation for EBV in SIP.

Acknowledgments

We would like to thank Professor Michael Haney for text revision, and PhD Anna Lindam for statistical advice.

Disclosure statement

The authors report no conflict of interest. The authors are alone responsible for the content and writing of the paper.

Additional information

Funding

This investigation was supported by grants from Lion’s Cancer Research Foundation Northern Sweden, Västerbotten County Council, Strategic Funds Clinical Science, Umeå University, Jämtlands Läns cancer och omvårdnadsfond and The Acta Oto-Laryngologica Foundation, Sweden.