The role of age and Merkel cell polyomavirus in oral cavity cancers

Abstract

The incidence of oral tongue cancer, the majority subsite of oral cavity cancer, is rising among young people with less exposure to tobacco and alcohol. Viral causes have been proposed, including Merkel cell polyomavirus (MCPyV). We evaluated patient and tumor characteristics among 126 incident oral cavity cancers (OCCs). Consistent with generational norms, younger patients had less exposure to tobacco and a greater number of oral sexual partners than older OCCs. In addition, younger patients were more likely to present at an earlier stage and with cancer arising from the oral tongue (each p<0.05). A subset of 44 cases were centrally tested for MCPyV Large T antigen expression by immunohistochemistry. In the presence of controls, none of the tumors expressed MCPyV. These findings exclude consideration of MCPyV as an etiologic factor of OCC, and may generate hypotheses for future examinations of the factors underlying the rise in oral tongue cancers.

Introduction

The incidence of oral tongue squamous cell carcinomas (OTSCCs), the largest subset of oral cavity cancers (OCCs), is rising in many countries.¹ This rise is seen most dramatically among young adult patients with less exposure to tobacco and alcohol.^2,3 Anecdotal experience suggests these OTSCCs characteristically arises on the lateral tongue with an increased rate of recurrence, although recent studies examining prognosis by age have produced heterogeneous results.^4,5 Interestingly, the lower burden of TP53 mutations among these cases is potentially suggestive of a viral etiology.⁶ Studies show OTSCC are not caused by HPV.^6,7 Indeed, to date, no attributable etiologic factor has been proven.

A recent PCR-based study of 21 OTSCCs arising in low-risk groups detected Merkel cell polyomavirus (MCPyV) DNA in 28% of cases.⁸ This analysis aimed to describe the characteristics of OCC by patient age, as well as to determine whether MCPyV is transcriptionally active in OCC.

Methods

Patients were enrolled in a prospective cohort study of head and neck squamous cell carcinoma (HNSCC) at three NCCN-designated Comprehensive Cancer Centers: The Sidney Kimmel Comprehensive Cancer Center at Johns Hopkins Hospital (JHH, Baltimore, MD), University of California San Francisco Helen Diller Family Comprehensive Cancer Center (UCSF, San Francisco, CA) and Tisch Cancer Institute at Icahn School of Medicine at Mount Sinai (New York, NY). This analysis was restricted to the 126 cases with oral cavity cancer (OCC). Demographic and behavioral data were collected using self-interview and medical record abstraction. Tumor subsite was collected by medical record review. Overlapping sites including tongue were considered tongue subsite. Categorical variables were compared by chi-squared methods. This study was approved by the Johns Hopkins Institutional Review Board, the UCSF Institutional Review Board, and the Icahn School of Medicine Institutional Review Board.

Tumor samples and staining:

A tissue microarray was constructed using JHH OCC samples with available formalin-fixed, paraffin-embedded (FFPE) tissue blocks. Antigen retrieval was performed on deparaffinized, 4 micron sections using 10 mM citrate buffer at 92°C for 30 minutes. A mouse monoclonal antibody for MCPyV Large T-antigen (Santa Cruz Biotechnology, Dallas, TX; clone CM2B4; 1:500 dilution) was applied using a BenchMark Ultra autostainer (Ventana Medical Systems, Tucson, AZ). Signals were visualized using the UltraView polymer detection kit (Ventana). Staining was performed according to manufacturer’s instructions with a known positive case of Merkel cell carcinoma as a positive control and benign tonsillar tissue as a negative control. The presence of staining was assessed by a head and neck pathologist (LR).

Human papillomavirus (HPV) testing was performed on evaluable whole-slide specimens from all sites using p16 immunohistochemistry and HPV RNA in situ hybridization (ISH) for high-risk types as described elsewhere⁹.

Results

Patient and tumor characteristics were compared by age group (Supplemental Table 1). Those 60 years or older (n=68, 54%) were similar to participants younger than 60 years (n=58, 46%) in terms of sex, race/ethnicity, marital status, income and education. However, compared to older patients, younger cases were significantly less likely to be ever smokers (42% vs. 67%, p=0.007), and more likely to have used illicit drugs (52% vs. 19%, p<0.001) and to have had 5 or more oral sex partners (52% vs. 19%, p<0.001). Fewer younger patients reported fair/poor oral hygiene (21% vs. 45%, p=0.01), diabetes (9% vs. 26%, p=0.02) and heart disease (8% vs. 25%, p=0.02). Lastly, younger patients in our cohort presented at an earlier AJCC 7^th ed. T and N stage. There were no differences in occupational exposure. Younger patients were more likely to present with oral tongue cancer (74% vs. 43%, p<0.001, Figure 1A).

Figure 1A.

Distribution of cancer subsite among 126 oral cavity cases, by age group.

Figure 1B. Proportion of 72 oral tongue cancer cases with lateral tongue subsite, by age group.

Characteristics were explored in a subset of 72 OTSCC. Similar differences in substance use, oral sexual exposure and comorbidity were seen by age. Among those with OTSCC, younger patients were as likely as older to present on the lateral oral tongue (p=0.16, Figure 1B). In addition, no differences in subsite were found when stratified by those with exposure to alcohol and/or tobacco and non-smoker/non-drinkers (p=0.39).

Forty-four OCC specimens were included on the TMA, of which 24 (55%) were OTSCC. All tumors were negative for MCPyV in the presence of positive and negative controls (Figure 2).

Figure 2A.

Oral squamous cell carcinoma (OCC) with hematoxylin and eosin stain (200x).

Figure 2B. Negative MCPyV Large T antigen staining among OCC (200x).

Figure 2C. Control case of MCPyV-positive Merkel cell carcinoma (400x).

HPV testing was performed on 106 specimens. Seventeen (15%) were p16-positive, and 9 (9%) were HPV positive by ISH. All ISH-positive tumors were p16-positive.

Discussion:

There is no clear etiology for the increasing incidence of oral tongue cancers. As such, the exploration of previously suggested factors including age, subsite and viral causes is highly important. Notable differences in behavioral risk factors for OCC were found by age. Though likely a reflection of generational norms rather than causative risk factors, these findings corroborate prior analyses.^12–14 In addition, we present a challenge to the notion¹⁵ that younger patients are more likely to present on the lateral tongue.

This is the first study designed to explore MCPyV transcriptional activity in OCC. We used immunohistochemistry to detect transcriptional activity and localize expression of MCPyV Large T-antigen,¹⁰ and found that MCPyV does not have a biologically meaningful role in OCC. Prior PCR-based studies reporting expression of MCPyV in OCC likely amplified viral copies ubiquitously found in saliva.^8,11 MCPyV is not responsible for the increasing incidence of oral tongue cancer.

Though limited by a small sample size, we present prospectively collected data describing incident OCC by age that may prompt further exploration in larger cohorts. Our results exclude further consideration of MCPyV as a causative factor in the rise of oral tongue cancer.

Supplementary Material

Supplemental Table.

Supplemental Table 1. Health history and tumor characteristics of patients with oral cavity cancer by age group.