Sage Journals: Discover world-class research

Abstract

A distinct clinical subgroup of nonsmoking (NS) and nondrinking (ND) patients with oral squamous cell carcinoma (OSCC) has been identified. The objective of the study was to assess this cohort for molecular variations in the disease process and if these could be attributed to clinical or epidemiological characteristics. One hundred and twenty-nine consecutive patients (71 males, 58 females) treated for OSCC were assessed at the Royal Melbourne Hospital between January 2007 and July 2010. Formalin-fixed paraffin embedded (FFPE) sections were stained for p53, p16, cyclin D1, and epidermal growth factor receptor (EGFR). Biomarker overexpression was observed in 72 (56%) cases for p53, 23 (18%) for p16, 45 (35%) for cyclin D1, and 72 (56%) for EGFR. Multiple logistic regression analysis revealed that tongue tumors (p = 0.012) and late stage cancers (p = 0.031) were more likely to have cyclin D1 overexpression. Further, older patients significantly more often had cyclin D1 overexpression (p = 0.008) and NSND patients had more p16 expression (p = 0.043). In contrast, smokers were more likely to have EGFR overexpression (p = 0.033). Concurrent overexpression of p53 and cyclin D1 were observed (p = 0.030). Smoking, site, and stage of OSCC can influence biomarker expression, with p16 overexpression specifically observed in NSND, indicating fundamental differences in the mechanisms of oral carcinogenesis among different patient cohorts.

Keywords

Immunohistochemistry oral squamous cell carcinoma biomarkers nondrinkers nonsmokers

Introduction

Oral squamous cell carcinoma (OSCC) is the sixth most prevalent malignancy globally.^1,2 Chronic tobacco use and alcohol consumption are well-established prominent risk factors.^3
–5 Oral cancer results in significant mortality and morbidity with severe impairment to quality of life, with the 5-year survival following diagnosis as low as 15–50% globally.^6,7 Fundamental to this, a large proportion of these cancers are diagnosed at an advanced stage and have associated lymphatic spread, explaining the poor outcomes.

Recently, a distinct clinical subgroup of elderly female nonsmoking (NS) and nondrinking (ND) patients with OSCC has been reported, which may represent an increasing proportion of OSCC patients over time due to declines in tobacco use and aging of the population.^8
–10 These investigators showed that this cohort had a different demographic distribution with worse outcomes. Currently, NSND population represents 13–35% of the OSCC population, with up to 24% of head and neck cancers being found in patients greater than 70 years of age.^11

–14

Over the last decade there has been much interest in molecular staging in cancer in a bid to improve patient management. Identification of relevant biomarkers may help in our understanding of oral carcinogenesis, have the potential to facilitate early diagnosis/detection of primary tumors, aid in the development of conservative therapies, and help in the prediction of disease progression and/or aggressiveness.^15,16

Abnormal cell cycle regulation is one of the hallmarks of carcinogenesis; dysregulation not only leads to uncontrolled cellular proliferation but also contributes to escape from other checkpoints such as senescence and apoptosis.¹⁷ Alterations in structure, function, or expression levels of proteins involved in cell cycle regulation are regarded to be important in the development and progression of head and neck cancer.¹⁸ The tumor suppressor protein p53 plays a pivotal role in cell cycle progression, with loss of function strongly associated with malignancy.^19,20 Cyclin D1 is also involved in cell cycle regulation (normal and neoplastic), allowing cells to progress from G1/S phase allowing the initiation of DNA replication.²¹ P16 inhibits progression from G1/S phase via the inhibition of cyclin D1-mediated phosphorylation of retinoblastoma protein.^18,22 The retinoblastoma pathway is crucial in regulating cell cycle progression; phosphorylation of retinoblastoma protein results in its functional inhibition and release of transcription factors required for cell cycle progression.²³

Epidermal growth factor receptor (EGFR) is a cell surface receptor for members of the family of epidermal growth factors, which activate tyrosine kinase signaling pathways resulting in cell proliferation, differentiation, and survival²⁴; overexpression has been described in various human malignancies.²⁵

This study’s hypothesis was that OSCC in NSND have a unique molecular profile and that this is reflected in its immunohistochemistry (IHC) expression to a panel of biomarkers. A panel of target proteins (biomarkers) as described above (p53, p16, cyclin D1, and EGFR) was chosen as they have been previously shown to have a high expression in oral cancer.²⁶

Methods

Patients

A previous study assessed 169 (97 males, 72 females) consecutive patients who presented with new or recurrent OSCC at the Royal Melbourne Hospital between January 2007 and July 2010.⁹ Of the 169 patients diagnosed, 11 had no tissue available as their diagnostic biopsy was undertaken prior to admission. This project was approved by the Melbourne Health Human Research Ethics Committee on 19th September 2012 (MH Project number 2012.071).

Tissue samples

FFPE tumor tissue blocks were obtained for each of the 158 patients. A pathologist (author CA) identified the area that contained the most tumor tissues. A total of 24 specimens were further excluded from the project as 18 had no discernable cancer in the surgical specimen and 6 had insufficient tissues. Thus, a total of 134 specimens of the original 169 were included for IHC analysis. Sections were cut from each identified block for IHC staining. However, following IHC, five specimens were further excluded as no tumor tissues were observable. Thus, this study was a convenience sample of 129 consecutive patients who had a diagnosis of OSCC and well-characterized clinical features, including history and known etiological agents. The study group in question was NSND patients (n = 32) with other OSCC patients (n = 97) as controls.

IHC

IHC staining was performed to determine the expression of the chosen biomarkers (p53, p16, cyclin D1, and EGFR). Formalin-fixed paraffin-embedded tissues were sectioned at 4 µm thickness and sequentially dewaxed/deparaffinized through a series of xylene, graded alcohol, and water immersion steps and then rehydrated and treated with 3% hydrogen peroxide to block endogenous peroxidase. Antigen retrieval was performed for all antibodies using Leica Bond high pH retrieval solution (Cat. no. AR9640) for 20 min at 100°C with the Leica Bond III stainer (Leica Microsystems, Inc., Buffalo Grove, Illinois, USA). The sections were mounted onto superfrost plus slides and air-dried at 37°C overnight.

IHC staining was performed on the Leica Bond III automated IHC stainer (using Leica Bond detection kit Cat. no. DS9800), according to the manufacturers’ instructions. The sections were equilibrated with neutralizing buffer (Leica wash buffer) and incubated for 15 min with primary antibodies. The primary antibodies and the dilutions used are shown in Table 1. After incubation with primary antibodies for 15 min at room temperature, bound antibodies were detected using the Bond polymer refine detection system, a biotin-free, polymeric horseradish peroxidase–linker antibody conjugate system (Leica Microsystems), which contains a peroxide block, post primary, polymer reagent, 3,3′-diaminobenzidine tetrahydrochloride chromogen, and hematoxylin counterstain.

Table 1.

Antibodies used and staining patterns observed.

Antigen	Primary antibody	Dilution	Staining pattern
p53	Mouse anti-p53 (Novocastra, North West England, UK)	1:50	Nuclear
p16	Mouse anti-p16 (Roche, New Jersey, USA)	1:3	Nuclear
Cyclin D1	Rabbit anti-cyclin D1 (Neomarkers, California, USA)	1:20	Nuclear and/or weak cytoplasm
EGFR	Mouse anti-EGFR (Novocastra, North West England, UK)	1:80	Cell membrane and/or weak cytoplasm

EGFR: epidermal growth factor receptor.

Positive control slides were included for validation in order to minimize variation in each IHC batch run using tissue sections that were known to be positive or had an abundance of the specific antibody.

Quantitative IHC analysis

All slides were scanned at an absolute magnification of 400× (resolution of 0.25 μm/pixel (100,000 pix/in.)) using the Aperio ScanScope XT system (Aperio Technologies Inc., Vista, California, USA). Slides were viewed and analyzed using ImageScope^® software package version 11.1.2.760 (Aperio Technologies Inc.) to quantify the IHC staining. All slides were coded and examined blindly. Three to five fields of view of representative areas of viable tumor of approximately the same size (0.025 mm²) were chosen for analysis under high-power fields. The degree of staining for p53, p16, and cyclin D1 was assessed using the positive pixel count algorithm, and EGFR assessed using the membrane algorithm. This latter method was used for EGFR staining as this is predominantly localized to the cell membrane, whereas the other proteins were predominately localized to the nucleus. As the membrane algorithm was used to evaluate EGFR staining, we could not analyze the extent of cytoplasmic staining that was weak in the majority of our cases. To confirm the reproducibility, 25% of the slides were randomly chosen and scored twice. All duplicates were evaluated using the methods described. Figure 1 shows a representation for overexpression of each biomarker.

Figure 1.

Representative immunostaining for overexpression of (a) p53, (b) p16, (c) cyclin D1, and (d) EGFR (magnification ×200). EGFR: epidermal growth factor receptor.

Statistical analysis

Raw data were initially collated into a Microsoft Excel spreadsheet and subsequently exported for all statistical analysis to the software package R (R Foundation for Statistical Computing, Vienna, Austria). Descriptive statistics were calculated for all variables.

For statistical analysis, the average positive pixel count result for each specimen was dichotomized and assigned one of two categories: negative/weak expression for scores of <20% and overexpression for scores >20% for p53, p16, and cyclin D1. For EGFR, the overall output score was categorized with scores of 0–2 assigned as negative/weak expression and a score of 3 assigned as overexpression (complete membrane staining).

The stage was classified according the American Joint Committee on Cancer sixth edition system and dichotomized for statistical analysis, with stages I and II classified as “early stage” and stages III and IV classified as “late stage.” Mortality data were sourced from both the ACCORD and the Victorian Cancer Registry databases, with a minimum follow-up of 18 months (census date January 1, 2012), including disease-specific mortality. Disease-free survival was recorded from the clinical follow-up notes.

Univariate analysis was performed using the χ ² test to assess the relationship between the epidemiological and clinical factors for each biomarker overexpression. Multivariate logistic regression analysis was then undertaken to identify the best set of predictors by estimation of the odds ratio (OR) and 95% confidence interval. Survival data (disease-free survival) for each biomarker was performed univariately with the Kaplan–Meier product limit method. The log-rank test was used to compare survival outcomes for statistical analysis. A p value of ≤ 0.05 was considered as significant.

Results

Patient characteristics

The study population consisted of 71 (55%) males and 58 (45%) females; 79 (61%) were less than 70 years of age and 50 (39%) were 70 years or older; 73 (57%) patients were smokers and 56 (43%) were nonsmokers (which included those with <5 pack years); 84 (65%) consumed alcohol (which included <3 drinks per week), with 45 (35%) being nondrinkers. The site was dichotomized and classified as tongue (comprising only cases from the anterior one-third or anterior to the foliate papillae) or nontongue (nontongue comprising mandibular and maxillary alveolus, retromolar region, floor and vestibule of mouth, hard palate, and cheek mucosa), with 49 (38%) of tumors originating from the tongue and the remaining 80 (62%) from nontongue; 72 (56%) patients had “early stage” OSCC, and those with “late stage” comprised 57 (44%) patients. A total of 38 (29%) patients’ deaths were due to OSCC. Table 2 provides a detailed summary of the clinicopathological characteristics analyzed. All tumors originated from the oral cavity (none were from the oropharyngeal region). No patients in this cohort were diagnosed with proliferative verrucous leukoplakia.

Table 2.

Demographics and clinical characteristics of patients with OSCC.

	Number of patients (%)
Gender
Male	71 (55)
Female	58 (45)
Age category
<70-year-old	79 (61)
>70-year-old	50 (39)
Tobacco
Smoker (S)	73 (57)
Nonsmoker (NS)	56 (43)
Alcohol
Drinker (D)	84 (65)
Nondrinker (ND)	45 (35)
Stage
I	34 (26)
II	38 (30)
III	23 (18)
IV	34 (26)
Site
Tongue (anterior 1/3rd)	49 (38)
Mandibular alveolus	5 (4)
Maxillary alveolus	17 (13)
Floor of mouth	18 (14)
Hard palate	6 (5)
Cheek mucosa	9 (7)
Vestibule of mouth	1 (<1)
Retromolar region	24 (19)

OSCC: oral squamous cell carcinoma.

Statistical analysis of the demographics of these patients showed that there were statistically significantly more males who consumed alcohol (χ ² = 8.6, df = 1, p value = 0.003) and were smokers (χ ² = 12.3, df = 1, p value = <0.001) compared to females. There were statistically significant more patients under the age of 70 years who consumed alcohol (χ ² = 6.2, df = 1, p value = 0.013) and were smokers (χ ² = 7.1, df = 1, p value = 0.008) compared to patients over the age of 70 years; but with death caused by OSCC occurring more in patients over the age of 70 years (χ ² = 10.8, df = 1, p value = 0.001). Later stage cancers were those of tongue origin (χ ² = 10.0, df = 1, p value = 0.002), with smokers also having significantly more tongue tumors (χ ² = 4.4, df = 1, p value = 0.036). Smoking was significantly associated with alcohol consumption, as expected (χ ² = 21.6, df = 1, p value ≤ 0.001).

Comparisons between the NSND group (n = 32 (25%)) and all other patients (n = 97 (75%)) found that the NSND cohort had statistically significant more females (χ ² = 18.9, df = 1, p value ≤ 0.001), a significantly greater proportion of individuals over the age of 70 years (χ ² = 12.9, df = 1, p value = 0.001) who had more tongue tumors (χ ² = 4.1, df = 1, p value = 0.042).

These results indicated that our cohort (n = 129) was representative of the initial cohort (n = 169) in that there were more females over the age of 70 years with tongue tumors, with a greater proportion of NSND as described previously.⁹

Biomarker expression

Overexpression of the biomarkers investigated was seen in 72 (56%) patients for p53, 23 (18%) for p16, 45 (35%) for cyclin D1, and 72 (56%) for EGFR. Tables 3 and 4 show the correlation of biomarkers with clinicopathological parameters. Patients who were less than 70 years of age had overexpression of cyclin D1 (χ ² = 10.2, df = 1, p = 0.001). Patients with later stage OSCC more often had higher levels of cyclin D1 and patients with tongue tumors also had higher levels of cyclin D1 overexpression, however, these differences did not reach statistical significance (χ ² = 3.6, df = 1, p value = 0.057; χ ² = 3.5, df = 1, p value = 0.062, respectively). Multiple logistic regression was undertaken to assess all epidemiological and clinical factors (gender, age, smoking status, drinking status, site, and stage of OSCC) for important predictor variables of biomarker overexpression to enable quantitative comparison of their separate and joint effects as indicated by the OR. Logistic regression was performed for each biomarker using backward stepwise regression after testing for any possible interactions among independent variables to eliminate the influence of confounders, with the best fit model used to determine the risk factors for each biomarker expression. Two separate models were assessed, the first for all patients and the second assessing the NSND cohort. In the model assessing all patients, no clinical variables were statistically significantly associated with the enhanced expression of p53 or p16. However, overexpression of cyclin D1 was statistically significantly associated with patients less than 70 years of age (OR = 3.67, 95% confidence interval (CI) = 1.57–9.30, p = 0.008), later stage cancers (OR = 2.57, 95% CI = 1.11–6.25, p = 0.031), and tongue tumors (OR = 3.03, 95% CI = 1.29–7.47, p = 0.013). The model assessing EGFR showed the overexpression of EGFR was statistically significantly higher in smokers (OR = 2.32, 95% CI = 1.09–5.14, p = 0.033). Logistic regression analysis of the NSND cohort alone showed increased p16 overexpression in those patients less than 70 years (OR = 11.43, CI = 1.40–245.23, p = 0.043).

Table 3.

Correlation of biomarkers with clinicopathological parameters (n = 129).^a

	p53		p16		Cyclin D1		EGFR
	Negative	Positive	Negative	Positive	Negative	Positive	Negative	Positive
Total	57	72	106	23	84	45	57	72
Gender
Male	29	42	58	13	43	28	32	39
Female	28	30	48	10	41	17	25	33
Age
<70-year-old	34	45	62	17	43	36	38	41
>70-year-old	23	27	44	6	41	9	19	31
Alcohol
Drinker (D)	38	46	70	14	52	32	39	45
Nondrinker (ND)	19	26	36	9	32	13	18	27
Tobacco
Smoker (S)	34	39	57	16	47	26	28	45
Nonsmoker (NS)	23	33	49	7	37	19	29	27
Site
Tongue	19	30	37	12	27	22	18	31
Nontongue	38	42	69	11	57	23	39	41
Stage
Early	31	41	59	13	52	20	32	40
Late	26	31	47	10	32	25	25	32
Death from oral cancer
No	42	49	73	18	62	29	43	48
Yes	15	23	33	5	22	16	14	24

EGFR: epidermal growth factor receptor.

^a“Negative” refers to no/weak expression and “positive” refers to overexpression.

Table 4.

Correlation of biomarkers with clinicopathological parameters in NSND cohort (n = 32).^a

	p53		p16		Cyclin D1		EGFR
	Negative	Positive	Negative	Positive	Negative	Positive	Negative	Positive
Total	12	20	27	5	22	10	13	19
Gender
Male	2	5	6	1	5	2	4	3
Female	10	15	21	4	17	8	9	16
Age
<70-year-old	5	6	7	4	6	5	5	6
>70-year-old	7	14	20	1	16	5	8	13
Site
Tongue	6	11	13	4	11	6	7	10
Nontongue	6	9	14	1	11	4	6	9
Stage
Early	6	12	16	2	15	3	8	10
Late	6	8	11	3	7	7	5	9
Death from oral cancer
No	9	10	15	4	15	4	8	11
Yes	3	10	12	1	7	6	5	8

EGFR: epidermal growth factor receptor; NS: nonsmoking; ND: nondrinking.

^a“Negative” refers to no/weak expression and “positive” refers to overexpression.

Results assessing the interactions of cyclin D1 with younger (<70 years), late stage, and tongue tumors showed that patients with late stage and tongue tumors (n = 10) were 9.9× more likely to have cyclin D1 overexpression (OR = 9.93, CI = 2.58–51.07, p = 0.002). Patients who were younger with late stage tongue tumors (n = 7) were 15.2× more likely to have cyclin D1 expression (OR = 15.23, CI = 2.59–291.08, p = 0.012). These patients also trended toward better survival compared to all other patients with cyclin D1 overexpression, however, this was not statistically significant (Figure 2(d)).

Figure 2.

Kaplan–Meier curves. “Negative” refers to no/weak expression and “positive” refers to overexpression. (a) NSND cohort versus others. (b) p16 overexpression versus p16 negative cohort. (c) NSND p16 negative cohort versus others. (d) Cyclin D1 overexpression cohort. NS: nonsmoking; ND: nondrinking.

Finally, multiple logistic regression was performed to identify correlations between biomarkers. From this it was found that p53 and cyclin D1 overexpression was significantly associated (OR = 2.32, CI = 1.10–5.09, p = 0.031).

The mean follow-up time for all patients was 28 months (range:18–64 months). Thirty-eight patients had died from oral cancer by the census date due to oral cancer. Kaplan–Meier curves were plotted from survival data. Analysis of all biomarker expression (log rank test) showed that none were statistically significantly associated with survival. The NSND cohort had a higher disease-specific mortality than the rest of the patient cohort, but this did not reach significance (p = 0.166; Figure 2(a)). However, poorer survival was seen in patients with p16 overexpression (Figure 2(b)); while in those that lacked p16 overexpression, the NSND cohort had lower survival (Figure 2(c)), however, neither reached significance (p = 0.084; p = 0.089, respectively).

The NSND cohort, consisting of 32 patients, also showed an increase in p16 overexpression in younger patients. Post hoc analysis was undertaken on the patients who had weak/no expression of p16, as it was felt that these individuals might be confounding the remaining analyses. Patients who were less than 70 years of age and those with tongue tumors had overexpression of cyclin D1 (χ ² = 7.5, df = 1, p = 0.006; χ ² = 6.3, df = 1, p = 0.012, respectively). NSND patients also had an overexpression of cyclin D1 (χ ² = 10.7, df = 1, p value = 0.001) but weak/no expression of p53, however, this difference did not reach statistical significance (χ ² = 3.6, df = 1, p value = 0.058).

Logistic regression further showed that cyclin D1 overexpression occurred statistically significantly more commonly in patients less than 70 years of age (OR = 3.94, 95% CI = 1.57–10.92, p = 0.005) and in those with tongue tumors (OR = 3.33, 95% CI = 1.37–8.42, p = 0.009). The model assessing EGFR showed that EGFR overexpression was statistically significantly more common in smokers (OR = 2.41, 95% CI = 1.05–5.73, p = 0.041).

Discussion

Koo et al.⁹ have demonstrated that NSND patients with OSCC comprise significantly more females over the age of 70, with the tongue as the predominant site of tumors that show poorer outcomes. Oral cancer has been traditionally found in males who smoke and drink, with the ratio of male to females in Australia previously reported as 2.5–3:1.²⁷ The cohort in the present study showed a male to female ratio to be 1.3:1. The aim of the present study was to assess this cohort for molecular variations in the disease process and if these could be attributed to clinical or epidemiological characteristics. Early detection remains to be the most promising approach to improve the long-term survival of patients with OSCC; biomarkers may provide important prognostic information for the management of patients with OSCC.

Overexpression of p53 was found in 56% of OSCC in the present study. This was consistent with the study by Bosari and Viale²⁸ who found that p53 expression in OSCC ranges from 37% to 78%. Furthermore, they reported that there was no correlation with clinicopathological variables, which is in agreement with our present study. Carlos de Vicente et al.²⁹ also showed inconclusive results when investigating if p53 immunoexpression influenced the clinical outcome of OSCC. They concluded that the ambiguity of results in the literature was attributed to a variety of factors including small and heterogeneous patient samples, tissue type analyzed (frozen and paraffin- embedded) and its pretreatment, antibodies applied, and the arbitrary threshold of the percentage of stained cells used to characterize sections as positive.²⁹

Cyclin D1 overexpression was found in 35% of patients in this study, which is similar to Shintani et al.³⁰ finding detection rates of 35.9%, with overexpression ranging between 39% and 83%. Later stage cancers were significantly more likely to have overexpression compared to earlier stages (p = 0.037) after adjusting for all other clinicopathological parameters in multivariate logistic analysis. This is in agreement with Sathyan et al.³¹ and Huang et al.³² who found that cyclin D1 overexpression was associated with stage, but not with the age, of onset. Other studies have shown no significant correlation between cyclin D1 overexpression and clinicopathological factors.^33,34 However, we found that patients ≥70 years of age were less likely to have cyclin D1 overexpression compared with those <70 years of age (p = 0.004). This may suggest that there are different mechanisms involved in tumor progression between younger and older patients.

The site of the tumor was also a predictor of cyclin D1 overexpression, with tongue tumors having an increased risk compared to other sites (p = 0.012), supporting the findings by Haas et al.³⁵ This finding suggests that OSCC originating in different sites of the oral cavity may behave differently. Other clinicoepidemiological studies show tongue cancer behavior is different from other sites of origin, but reasons for such difference are lacking in the literature.³¹ Our results indicate that OSCC of different anatomical subsites are actually characterized by alterations at different pathways.

EGFR was found in 56% of patients, a result aligned with previous reports.³⁶ Therapeutic agents targeting EGFR, including EGFR inhibitor cetuximab, have been developed with promising results,³⁷ however, clinical information concerning its status in OSCC is limited and varied. Ryott et al.³⁸ found no correlation with EGFR status and survival, but significant expression in advanced tumors (stages II, II, and IV), and that nonsmokers with tongue cancers had a higher expression in early stages (I and II) compared to smokers. However, Kumar et al.³⁹ found smoking was significantly associated with EGFR expression, and these patients with oropharyngeal cancers had the worst overall survival. This is supported by Moraitis et al.⁴⁰ who found that EGFR ligand levels are increased in the oral mucosa of smokers. These latter findings therefore are in accordance with the result of the present study in that smokers were found to have a statistically significant increased risk of EGFR overexpression (p = 0.032).

P16 overexpression was found in 18% of the patients, with 16% of the NSND cohort having p16 overexpression. p16 has been reported to show a significant correlation with head and neck cancers in nonsmokers, specifically for patients whose cancers originated in the tonsils.³⁵ In the present study, we found that in the NS and ND cohort there was a significant association between p16 overexpression in those <70 years. This may suggest that tumors in those <70 years are genetically distinct. Su et al.⁴¹ found that hypermethylated p16 promoter was associated with regional invasiveness, distant metastasis, and reduced life span in elderly patients, suggesting that the molecular carcinogenic mechanism of p16 might be related in part to aging. However, criticizing this study Shaw et al.⁴² have suggested that due to the small sample size and great risk of type I error, their results should not go unchallenged, as other studies have shown that p16 methylation was associated with OSCC presence but not poor prognosis.^43,44 The p16 IHC has been used as a surrogate marker for human papillomavirus (HPV) infection in oropharyngeal cancer.^45

–48 Duncan et al.¹ found that strong staining for p16 correlated with HPV positivity in OSCC. However, others have found no correlation between p16 overexpression and the presence of HPV.^3,5 HPV-associated cancers have been shown to be less likely to occur among drinkers and smokers,⁶ and further oral tongue cancers had no significant association with HPV.^8,10 The results of the present study indicate that there may be a cohort of younger, NSND patients with oral cancer caused by HPV. Further investigation such as polymerase chain reaction (PCR) or in-situ hybridization, would be needed to confirm the presence of HPV, as the evidence for the association between p16 overexpression and HPV infection is conflicting.

Analysis of survival rates showed there was a higher disease-specific mortality in the NSND cohort (Figure 2(a)) and p16 overexpression cohort (Figure 2(b)). Those who were nonsmokers and nondrinkers and were p16 negative had higher disease-specific mortality (Figure 2(c)). This finding is contrary to previous reports, which have reported that those with p16 overexpression tend to have better outcomes and overall survival.^11,13,14 However, these studies assessed both oral cavity and oropharyngeal squamous cell carcinoma, and it may well be that patients with p16-positive oral cavity tumors have a far worse progression than p16-positive oropharyngeal cancer patients.

The results of the present study indicate that p53 and cyclin D1 overexpression are significantly correlated, which is in accordance with previous studies.^13,34 This may suggest that loss or dysregulation of cell cycle control plays a pivotal role in oral carcinogenesis due to complex interactions between the proteins involved in cell cycle regulation.

Previous reports⁴⁹ have concluded that clinical and genetic features of head and neck squamous cell carcinoma are distinct between smokers and nonsmokers, with tumors of nonsmokers containing a lower frequency of common genetic alterations, but the cause of these underlying changes still remains unknown.⁵⁰ Focusing on the NSND population with OSCC may provide clues of the path to tumorigenesis, as it is independent of the effect of known carcinogens.

The findings from this study indicate distinct biomarker overexpression in OSCC occurs in different age-groups, tumor sites, and stages. Those with cyclin D1 overexpression appear to be a distinct clinical subgroup of patients who are younger, have later stage cancers presenting in the tongue. Patients with all three of these clinical features (n = 7) were 15 times more likely to have cyclin D1 overexpression. There was no significant statistical association with survival; however, these patients did trend toward better survival, when compared to all other patients with cyclin D1 overexpression (Figure 2(d)). A link between young NSND patients and the overexpression of p16 in their tumors has also been demonstrated; however, further investigations are still needed to correlate this biomarker overexpression with clinical attributes of disease and the association with HPV. In particular, it is now necessary to undertake quantitative real-time PCR to corroborate the IHC findings for each biomarker in the present study and, most importantly, assess the significance of our p16 IHC results.

Footnotes

Acknowledgments

The authors acknowledge the financial assistance of the Price Family Foundation for this project, and BioGrid for assistance with the ACCORD Head and Neck Database.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Duncan

Winkler

Carlson

. p16 immunohistochemistry can be used to detect human papillomavirus in oral cavity squamous cell carcinoma. J Oral Maxillofac Surg 2013; 71: 1367–1375.

Warnakulasuriya

. Global epidemiology of oral and oropharyngeal cancer. Oral Oncol 2009; 45: 309–316.

Singhi

Westra

. Comparison of human papillomavirus in situ hybridization and p16 immunohistochemistry in the detection of human papillomavirus-associated head and neck cancer based on a prospective clinical experience. Cancer 2010; 116: 2166–2173.

Hashibe

Brennan

Chuang

S-C

. Interaction between tobacco and alcohol use and the risk of head and neck cancer: pooled analysis in the International Head and Neck Cancer Epidemiology Consortium. Cancer Epidemiol Biomarkers Prev 2009; 18: 541–550.

Poling

Bui

. Human papillomavirus (HPV) status of non-tobacco related squamous cell carcinomas of the lateral tongue. Oral Oncol 2014; 50: 306–310.

Gillison

Koch

Capone

. Evidence for a causal association between human papillomavirus and a subset of head and neck cancers. J Natl Cancer Inst 2000; 92: 709–720.

McCullough

Farah

. The role of alcohol in oral carcinogenesis with particular reference to alcohol-containing mouthwashes. Aust Dent J 2008; 53: 302–305.

Dahlgren

Dahlstrand

Lindquist

. Human papillomavirus is more common in base of tongue than in mobile tongue cancer and is a favorable prognostic factor in base of tongue cancer patients. Int J Cancer 2004; 112: 1015–1019.

Koo

Barrowman

McCullough

. Non-smoking non-drinking elderly females: a clinically distinct subgroup of oral squamous cell carcinoma patients. Int J Oral Maxillofac Surg 2013; 42: 929–933.

10.

Liang

X-H

Lewis

Foote

. Prevalence and significance of human papillomavirus in oral tongue cancer: the mayo clinic experience. J Oral Maxillofac Surg 2008; 66: 1875–1880.

11.

Lewis

Thorstad

Chernock

. p16 positive oropharyngeal squamous cell carcinoma: an entity with a favorable prognosis regardless of tumor HPV status. Am J Surg Pathol 2010; 34: 1088–1096.

12.

Kruse

Bredell

Luebbers

. Head and neck cancer in the elderly: a retrospective study over 10 years (1999-2008). Head Neck Oncol BioMed Central 2010; 2: 25.

13.

Shah

Trivedi

Tankshali

. Prognostic significance of molecular markers in oral squamous cell carcinoma: a multivariate analysis. Head Neck 2009; 31: 1544–1556.

14.

Reimers

Kasper

Weissenborn

. Combined analysis of HPV-DNA, p16 and EGFR expression to predict prognosis in oropharyngeal cancer. Int J Cancer 2007; 120: 1731–1738.

15.

Riley

Sauerbrei

Altman

. Prognostic markers in cancer: the evolution of evidence from single studies to meta-analysis, and beyond. Br J Cancer 2009; 100: 1219–1229.

16.

Polanska

Raudenska

Gumulec

. Clinical significance of head and neck squamous cell cancer biomarkers. Oral Oncol 2014; 50: 168–177.

17.

Lundberg

Weinberg

. Control of the cell cycle and apoptosis. Eur J Cancer 1999; 35: 1886–1894.

18.

Todd

Hinds

Munger

. Cell cycle dysregulation in oral cancer. Crit Rev Oral Biol Med 2002; 13: 51–61.

19.

Vousden

Lane

. p53 in health and disease. Nat Rev Mol Cell Biol 2007; 8: 275–283.

20.

de Oliveira

Ribeiro-Silva

Zucoloto

. Prognostic impact of p53 and p63 immunoexpression in oral squamous cell carcinoma. J Oral Pathol Med 2007; 36: 191–197.

21.

Donnellan

Chetty

. Cyclin D1 and human neoplasia. Mol Pathol 1998; 51: 1–7.

22.

Pande

Mathur

Shukla

. pRb and p16 protein alterations in human oral tumorigenesis. Oral Oncol 1998; 34: 396–403.

23.

Weinberg

. The retinoblastoma protein and cell cycle control. Cell 1995; 81: 323–330.

24.

Herbst

. Review of epidermal growth factor receptor biology. Int J Radiat Oncol Biol Phys 2004; 59: S21–S26.

25.

Rubin Grandis

Melhem

Barnes

. Quantitative immunohistochemical analysis of transforming growth factor-alpha and epidermal growth factor receptor in patients with squamous cell carcinoma of the head and neck. Cancer 1996; 78: 1284–1292.

26.

Oliveira

Ribeiro-Silva

. Prognostic significance of immunohistochemical biomarkers in oral squamous cell carcinoma. Int J Oral Maxillofac Surg 2011; 40: 298–307.

27.

Abreu

Kruger

Tennant

. Oral cancer in Western Australia, 1982-2006: a retrospective epidemiological study. J Oral Pathol Med 2010; 39: 376–381.

28.

Bosari

Viale

. The clinical significance of p53 aberrations in human tumours. Virchows Arch 1995; 427: 229–241.

29.

Carlos de Vicente

Junquera Gutiérrez

Zapatero

. Prognostic significance of p53 expression in oral squamous cell carcinoma without neck node metastases. Head Neck 2004; 26: 22–30.

30.

Shintani

Mihara

Nakahara

. Expression of cell cycle control proteins in normal epithelium, premalignant and malignant lesions of oral cavity. Oral Oncol 2002; 38: 235–243.

31.

Sathyan

Sailasree

Jayasurya

. Carcinoma of tongue and the buccal mucosa represent different biological subentities of the oral carcinoma. J Cancer Res Clin Oncol 2006; 132: 601–609.

32.

Huang

S-F

Cheng

S-D

Chuang

W-Y

. Cyclin D1 overexpression and poor clinical outcomes in Taiwanese oral cavity squamous cell carcinoma. World J Surg Oncol 2012; 10: 40.

33.

Kyomoto

Kumazawa

Toda

. Cyclin-D1-gene amplification is a more potent prognostic factor than its protein over-expression in human head-and-neck squamous-cell carcinoma. Int J Cancer 1997; 74: 576–581.

34.

Lam

Yuen

. Cyclin D1 expression in oral squamous cell carcinomas: clinicopathological relevance and correlation with p53 expression. J Oral Pathol Med 2000; 29: 167–172.

35.

Haas

Hörmann

Bosch

. Expression of cell cycle proteins in head and neck cancer correlates with tumor site rather than tobacco use. Oral Oncol 2002; 38: 618–623.

36.

Bei

Budillon

Masuelli

. Frequent overexpression of multiple ErbB receptors by head and neck squamous cell carcinoma contrasts with rare antibody immunity in patients. J Pathol 2004; 204: 317–325.

37.

Vermorken

Mesia

Rivera

. Platinum-based chemotherapy plus cetuximab in head and neck cancer. N Engl J Med 2008; 359: 1116–1127.

38.

Ryott

Wangsa

Heselmeyer-Haddad

. EGFR protein overexpression and gene copy number increases in oral tongue squamous cell carcinoma. Eur J Cancer 2009; 45: 1700–1708.

39.

Kumar

Cordell

Lee

. Response to therapy and outcomes in oropharyngeal cancer are associated with biomarkers including human papillomavirus, epidermal growth factor receptor, gender, and smoking. Int J Radiat Oncol Biol Phys 2007; 69: S109–S111.

40.

Moraitis

De Lorenzo

. Levels of cyclooxygenase-2 are increased in the oral mucosa of smokers: evidence for the role of epidermal growth factor receptor and its ligands. Cancer Res 2005; 65: 664–670.

41.

P-F

Huang

W-L

H-T

. p16INK4A promoter hypermethylation is associated with invasiveness and prognosis of oral squamous cell carcinoma in an age-dependent manner. Oral Oncol 2010; 46: 734–739.

42.

Shaw

Risk

Liloglou

. RE: P16 INK41 promoter hypermethylation is associated with invasiveness and prognosis of oral squamous cell carcinoma in an age dependent manner. Su et al. Oral Oncology 46 (2010) 734–739. Oral Oncol 2011; 47: 776.

43.

Hall

Shaw

Field

. p16 Promoter methylation is a potential predictor of malignant transformation in oral epithelial dysplasia. Cancer Epidemiol Biomarkers Prev 2008; 17: 2174–2179.

44.

Tan

Saulnier

Auperin

. Quantitative methylation analyses of resection margins predict local recurrences and disease-specific deaths in patients with head and neck squamous cell carcinomas. Br J Cancer 2008; 99: 357–363.

45.

Fregonesi

PAG

Teresa

Duarte

. p16INK4A immunohistochemical overexpression in premalignant and malignant oral lesions infected with human papillomavirus. J Histochem Cytochem 2003; 51: 1291–1297.

46.

Smeets

Hesselink

Speel

EJM

. A novel algorithm for reliable detection of human papillomavirus in paraffin embedded head and neck cancer specimen. Int J Cancer 2007; 121: 2465–2472.

47.

Marur

D’Souza

Westra

. HPV-associated head and neck cancer: a virus-related cancer epidemic. Lancet Oncol 2010; 11: 781–789.

48.

Benson

Eisele

. The clinical impact of HPV tumor status upon head and neck squamous cell carcinomas. Oral Oncol 2014; 50: 565–574.

49.

Vargas-Ferreira

Nedel

Etges

. Etiologic factors associated with oral squamous cell carcinoma in non-smokers and non-alcoholic drinkers: a brief approach. Braz Dent J 2012; 23: 586–590.

50.