Journal of International Oral Health

: 2019  |  Volume : 11  |  Issue : 2  |  Page : 92--99

Evaluation of tumor necrosis factor: Alpha in the saliva of oral cancer, leukoplakia, and healthy controls – A comparative study

M Ameena, R Rathy 
 Department of Oral Pathology, Azeezia College of Dental Sciences and Research, Kollam, Kerala, India

Correspondence Address:
M Ameena
Department of Oral Pathology, Azeezia College of Dental Sciences and Research, Kollam, Kerala


Aims: The survival of oral squamous-cell carcinoma (OSCC) patients remains poor despite recent treatment advances. A sensitive and specific biomarker is important in screening high-risk patients. The present study was undertaken to test a hypothesis whether salivary tumor necrosis factor-alpha (TNF-α) can be used as a biomarker for OSCC. The study aimed to assess salivary TNF-α in OSCC, leukoplakia, and whether it can be used as a biomarker for the early diagnosis of OSCC. The objectives are as follows: (1) To evaluate salivary TNF-α and compare with histological grades of OSCC. (2) To evaluate salivary TNF-α in leukoplakia and compare with different grades of dysplasia. (3) To compare TNF-α levels in the saliva of oral cancer patients with leukoplakia patients and with healthy control group. Materials and Methods: The study was conducted in 90 participants, of which 30 healthy individuals, 30 leukoplakia, and 30 OSCC patients. Whole unstimulated saliva was collected and analyzed using an ELISA test. Results: TNF-α was significantly elevated in leukoplakia, further elevated in OSCC as compared to controls. There was a significant difference in TNF-α between the different histopathological grades of OSCC and leukoplakia (P ≤ 0.01). There were also statistically significant differences in TNF-α level between different clinical stages in OSCC (P ≤ 0.05). ROC curve analysis and area under curve values showed high specificity and sensitivity in differentiating OSCC from leukoplakia and healthy controls. Conclusion: The present study shows that salivary TNF-α can be used as a marker for predicting leukoplakia and oral cancer. The study also showed a significant correlation between clinical staging and histopathological grading of OSCC and TNF-α level.

How to cite this article:
Ameena M, Rathy R. Evaluation of tumor necrosis factor: Alpha in the saliva of oral cancer, leukoplakia, and healthy controls – A comparative study.J Int Oral Health 2019;11:92-99

How to cite this URL:
Ameena M, Rathy R. Evaluation of tumor necrosis factor: Alpha in the saliva of oral cancer, leukoplakia, and healthy controls – A comparative study. J Int Oral Health [serial online] 2019 [cited 2019 Jul 19 ];11:92-99
Available from:

Full Text


Oral cancer accounts for 2%–4% of all cancer cases worldwide and >90% of all oral cancers are oral squamous-cell carcinoma (OSCC).[1] Despite the easy accessibility of oral cavity for examination, most oral cancers are detected at a later stage leading to lower survival rates.[2] The high morbidity can be attributed to the delay in the diagnosis.[3]

Tumor necrosis factor-alpha (TNF-α), an inflammatory mediator, has been implicated in carcinogenesis, due to its participation in chronic inflammatory diseases. Low, sustained production can induce a tumor phenotype.[4] It act as an endogenous tumor promoter to bridge inflammation and carcinogenesis. Based on the conducted studies, it could be used as an indicator of risk, therapy response, and prognosis for cancer patients.[5] Krishnan et al. observed salivary TNF-α in participants with premalignant lesion and OSCC among patients reported to the OPD of SRM dental college in Chennai. Salivary TNF-α was significantly higher than serum level in OSCC groups in their study. Most of the studies investigated TNF-α levels in saliva and serum without regard to oral inflammatory conditions and periodontal conditions which might be one of the reasons for wide variability in the values between diseased and healthy controls in various studies. The present study was undertaken to evaluate salivary TNF-α among patients with leukoplakia and OSCC and also to compare the salivary levels of TNF-α with histopathological grades of leukoplakia and OSCC.

 Materials and Methods

The quantitative analytical study was carried out from July 2014 to July 2015 (duration of the study is 1 year) in the patients availing treatment in the OPD of Azeezia Dental College and Azeezia Medical College during the above said period. The study was commenced on obtaining clearance from the Institutional Ethical Committee (ACDS/2158/1/2013). The sex-matched and age-matched 30 healthy participants were selected as control group. The participants were thoroughly examined and detailed case history was taken from each participant as per the attached pro forma. Written informed consents were obtained from all the participants.

Patients with OSCC and with leukoplakia were included in the study. Exclusion criteria includes participants with acute and chronic inflammatory conditions, systemic diseases, participants taking drugs that induce hyposalivation or hypersalivation such as anticholinergics, antihistamines, antihypertensives, and beta-blockers were excluded from the study. Patients undergoing treatment for oral cancer and leukoplakia and with other oral mucosal lesions were also excluded. Participants of study group with clinical attachment loss ≥4 mm in more than 14 teeth were excluded. If total number of teeth present in the subject was <14, 8–13 teeth were examined (Classification of Periodontal Diseases and Conditions by the American Academy of Periodontology 1999).[6]

A simple random sampling technique is performed. The study group study group included 30 participants with OSCC and 30 participants with leukoplakia. Out of 30 participants with leukoplakia, 15 were low-risk leukoplakia (no/questionable/mild dysplasia) and others were high-risk leukoplakia (moderate or severe dysplasia).

The sample size is to determine a mean difference (80% power, 0.05 significance two-sided test).


n = sample size σ = sp = pooled standard deviation Δ = expected mean difference.

The control group included 30 healthy participants who were age- and gender-matched with good oral health status, normal occlusion, and healthy periodontium.

Whole unstimulated saliva was collected by passive drool (Navazesh et al.). Saliva samples were collected before breakfast between 8 a.m. and 10 a.m. Participant lets the saliva passively drip into a custom-made saliva collecting funnel over a period of 5–10 min till approximately 1.5 mL of saliva was collected. Samples were stored at −20°C until analysis. The stored saliva samples were centrifuged at 4000 r.p.m for 10 min at 4°C. TNF-α was estimated using Cayman TNF-alpha (human) EIA kit (Item no. 589201) by solid-phase sandwich ELISA.

Statistical analysis was performed using the one-way ANOVA, Independent t-test, Scheffe multiple comparisons post hoc test, and ROC curve analysis.


Of 90 participants, 62 were male and 28 were female. There was no significant association found in gender of different groups [Table 1].{Table 1}

The mean age in OSCC, leukoplakia, and control groups were found to be 57.8 ± 9.4, 49.2 ± 10.1, and 57.2 ± 9.2 years, respectively. There was a significant difference in the distribution of participants in different age groups in the one-way ANOVA analysis [Table 2].{Table 2}

There was a significant difference in the habits between these two groups in Chi-square test. 93.3% of the OSCC group had chewing habit while subjects with chewing habit accounts for only 63.3% in leukoplakia group [Table 3] and [Graph 1]. No significant association of duration of habits between the two study groups [Table 4] and [Graph 2]. Buccal mucosa is the most common site in 40% of OSCC and 60% of leukoplakia cases [Table 5] and [Graph 3]. There was no statistically significant difference in average clinical attachment loss between the three groups in one-way ANOVA analysis [Table 6].{Table 3}[INLINE:2]{Table 4}[INLINE:3]{Table 5}[INLINE:4]{Table 6}

TNF-α was compared using ANOVA test. Scheffe multiple comparison analysis was used to compare TNF-α level between each of these groups. A statistically significant difference was found in TNF-α level between OSCC and leukoplakia, OSCC and controls, and leukoplakia and controls (P = 0.000) [Table 7] and [Graph 4]. In Independent t-test, a significant difference in TNF-α level between well-differentiated group and moderate/poorly differentiated group was observed [Table 8] and [Graph 5]. A significant difference in the TNF-α level based was observed in one-way ANOVA test (P = 0.30). Stages III and IV showed increased TNF–α level when compared to Stages I and II [Table 9] and [Graph 6]. No significant association between TNF-α level and size of the leukoplakia was observed in one-way ANOVA test [Table 10] and [Graph 7]. Independent t-test was used to test the significance of TNF-α level between histopathologic grades of leukoplakia. A significant difference between the different grades was observed (P = 0.000) [Table 11] and [Graph 8]. A sensitivity and specificity of 93.3% each was obtained in ROC curve analysis. Area under the curve was found to be 0.992 [Table 12] and [Graph 9].{Table 7}[INLINE:5]{Table 8}[INLINE:6]{Table 9}[INLINE:7]{Table 10}[INLINE:8]{Table 11}[INLINE:9]{Table 12}[INLINE:10]


Most OSCC are not diagnosed until an advanced stage, even though oral cavity is easily accessible for direct visual examination, which may be one of the major reasons for the low-survival rate. Salivary biomarkers gained the importance as collection of saliva is relatively easy and noninvasive, and hence, the search for reliable salivary biomarkers for the early detection of OSCC has developed rapidly.[7]

TNF-α has tumor-promoting role in various stages of carcinogenesis. It is related to RONS generation and promotes oxidative stress-mediated DNA damage. It stimulates TGF-α-induced epithelial-to-mesenchymal transition and also induces secretion of vascular endothelial growth factor by human fibroblasts and promoting angiogenesis. It plays a role in inducing NF-κB signaling, a decisive pathway in driving metastasis.[4] Cytokines like TNF-α with proinflammatory, proangiogenic, and immunoregulatory activity are produced by OSCC, become a part of the local tumor environment and contribute to the progression of oral cancer.[8] A statistically significant difference in the TNF-α levels of OSCC, leukoplakia, and control group (P = 0.000) was noted in the present study as in hypothesis.

A male proclivity was observed in the present study. Of 60 participants of the study group, 40 were male and only 20 were female. This can be attributed to the increased prevalence of tobacco use and alcoholism in males than in females.

The average age of patients with OSCC and leukoplakia in the present study was 57.8 and 49.2 years, respectively. This was in accordance with Juretić et al. (2013),[9] who reported in their study, an average age to be 55.1 and 54.2 years for OPML and OSCC patients, respectively. Brailo et al.[10] reported a mean age of 52.3 and 52 years in leukoplakia and healthy control groups, respectively, in their study.

In the descriptive study of four different clinical stages in oral squamous cell carcinoma it was found that, 40% of subjects were in stage III, followed by stage II, IV and I respectively [Table 13].{Table 13}

Well differentiated accounts for 73.3% of cases followed by 23.3% moderately differentiated and 3.3% poorly differentiated in the percentage distribution of histopathologic grades in Oral squamous cell carcinoma [Table 14].{Table 14}

The average size of leukoplakia was found to be 1.8 ± 1.1centimetres. 53.3% of cases were found to have size ranging between 1-2.5 cm [Table 15].{Table 15}

76.7% of leukoplakia cases were in low risk group while 23.3% was in high risk group [Table 16].{Table 16}

A significant difference was observed in the habits between the two study groups. In the present study, 93.3% of the OSCC patients and 63.3% of leukoplakia patients had chewing habit. All the patients in the study group had a minimum of one deleterious habit which was in accordance with the study by Krishnan et al.[11]

It was found that, 63.3% of the OSCC patients were alcohol users, whereas in the leukoplakia group, it was only 16.7%. Duration of the habits ranged between 26 and 35 years in 43% of OSCC patients and 40% of leukoplakia. A higher incidence of OSCC and leukoplakia has been reported with an increase in duration of deleterious habits.

The most common site was the buccal mucosa (40%), followed by alveolus and gingiva (20%), tongue (17%), vestibule (10%), retromolar (7%), commissure (3%), and floor of the mouth (3%) in patients with OSCC. In case of site distribution of lesions in patients with leukoplakia, the common site was the buccal mucosa (60%), followed by commissure (20%), gingiva (10%), and tongue (10%). This was similar to the findings of Krishnan et al.[11]

The exclusion criteria of the present study was comparable to the studies by SahebJamee et al.,[12] Brailo et al.,[13] Patil et al.,[14] Juretić et al.,[9] and Krishnan et al.[11]

The present study also excluded chronic generalized periodontitis by eliminating participants with pocket depth ≥4 mm in more than 14 teeth. If total number of teeth present in the participant were <14 in which 8–13 teeth were examined. The periodontal health between the three groups was comparable. One-way ANOVA analysis showed no statistically significant difference in the average clinical attachment loss between the three groups. Age- and sex-matched participants with healthy periodontium were included in the control group. SahebJamee et al.[12] used modified gingival index to assess gingival condition of their participants, whereas Brailo et al.[13] and Brailo et al.[10] used CPITN in their study. The present study emphasized exclusion of patients with periodontitis in the study group.

Salivary TNF-α concentration for the OSCC group was 133.3 ± 15.0 pg/mL, whereas for leukoplakia group, it was 99.8 ± 7.9 pg/mL, and 83.3 ± 5.5 pg/mL for control participants in the present study. A statistically significant difference in the TNF-α levels of OSCC, leukoplakia, and control group (P = 0.000) was also noted.

Krishnan et al.[11] observed salivary TNF-α in control as 4.5 ± 2.5 pg/mL, premalignant lesion participants as 136.8 ± 59 pg/mL, and OSCC as 311.9 ± 95.3 pg/mL using sandwich ELISA method (Syntron Bioresearch In). Salivary TNF-α is significantly higher than the serum level in OSCC groups in their study. ROC curve analysis in their study revealed salivary TNF-α as a better medium for detecting OSCC.

Patil et al.[14] reported a salivary TNF-α concentration of 112.68 ± 42.63 pg/mL, 28 ± 6.37 pg/mL, and 21.56 ± 53.61 pg/mL in 50 participants each with OSCC, leukoplakia, and controls, respectively. The difference between OSCC and leukoplakia, OSCC and controls (P = 0.018) were found to be statistically significant (P = 0.027), whereas a statistically nonsignificant (P = 0.243) relation was observed between leukoplakia and control group. TNF-α concentrations were determined by the quantitative sandwich ELISA technique using commercially available kits (Bender MedSystem, Austria).

A study by Juretić et al. (2013)[9] in 19 patients with oral premalignant lesions, 19 with OSCC, and 19 healthy controls observed a statistically significant (P < 0.001) differences in salivary concentration of TNF-α levels between all groups in the ELISA test. Participants with OSCC had the highest (0.739 ± 0.176 pg/ml), OPML (0.601 ± 0.178 pg/ml) and control group (0.013 ± 0.033 pg/ml) had the lowest concentration of TNF-α.

Rhodus et al.[8] observed a significant difference in TNF-α levels in three groups of 13 participants with OPML, OSCC, and controls by ELISA test. TNF-α levels were found to be OSSC = 28.9 ± 14.6 pg/mL, OPML = 10.5 ± 7.4 pg/mL, and controls 3.0 ± 1.0 pg/mL (P < 0.01).

Brailo et al.[10] observed significantly increased levels of salivary TNF-α level in 30 participants with leukoplakia (9.23 ± 10.6 pg/mL) than in 34 healthy participants (4.06 ± 7.46 pg/ml) by the ELISA test.

SahebJamee et al.[12] in a study with nine OSCC patients and nine controls found statistically insignificant relation of TNF-α levels. TNF-α levels were determined by the quantitative sandwich ELISA technique using commercially available kits and were found 35.2 ± 51.8 pg/mL for OSCC and 4.1 ± 2.1 pg/mL for controls (P > 0.05).

The present study was contradictory to that of Brailo et al.[13] who observed no differences in concentrations of salivary TNF-α between oral cancer (n = 28), leukoplakia (n = 29), and controls (n = 31). Salivary TNF-α levels were reported to be 34 ± 21.58 in patients with oral cancer, 30 ± 3.01 in leukoplakia, and 38 ± 3.23 pg/ml in controls (P = 0.126) in commercial chemiluminescent enzyme-linked immunoassay.

A significant difference in TNF-α level between well-differentiated OSCC group (126.9 ± 11.9 pg/ml) and moderate/poorly differentiated OSCC group (150.7 ± 6.4 pg/ml) was noted in the study similar to Krishnan et al.[11] Krishnan et al.[11] reported the median and interquartile range of salivary TNF-α levels as 281 (120.3–390.4) pg/mL, 293.9 (178.5–593.8) pg/mL, and 310.2 (195.7–544.4) pg/mL for well-differentiated, moderately differentiated, and poorly differentiated lesions.

A significant difference in the TNF–α level was observed between the different clinical stages. Stages III (133.8 ± 17.0 pg/ml) and IV (147.4 ± 14.3 pg/ml) showed increased TNF–α level when compared to Stages I (124.6 ± 4.3 pg/ml) and II (126.2 ± 7.5 pg/ml) (significant at 0.05 level) which was comparable to the study by Krishnan et al.[11] In their study, the median and interquartile range of TNF-α for Stage I was 270.1 pg/mL (120.3–218.7), Stage II was 289.3 pg/mL (1780.2–267.5), Stage III was 290.1 pg/mL (190.2–390.4), and Stage IV was 501.4 (390.1–593.8) pg/ml, respectively.

Kurokawa and et al.[15] observed a better prognosis in the Kaplan–Meier curve analysis for patients with serum TNF-alpha-positive than for patients with serum TNF-alpha-negative. Korostoff et al.[16] reported that TNF-alpha was elevated in the endophytic compared to exophytic tongue SCC, which correlated with the decreased survival rate in this group. The present study was a cross-sectional study and hence, the comparison of TNF-alpha with patient survival could not be undertaken.

The average size of leukoplakia lesion was found to be 1.8 ± 1.1 cm. Nearly 53.3% of cases were of size between 1 cm and 2.5 cm followed by 30% >2.5 cm and 16.7% <1 cm. No significant association was observed between TNF-α level and size of leukoplakia in one-way ANOVA test. This was comparable with the study by Brailo et al.[10]

The present study also compared TNF-α level between the histopathologic grades of epithelial dysplasia. A significant difference was observed between high-risk and low-risk groups. To the best of our knowledge, only few studies have been reported comparing the TNF-α level between histopathologic grades of epithelial dysplasia. Rhodus et al.[17] reported that TNF-α, interleukin 1 (IL-1), IL-6, and IL-8 were significantly higher in moderate-and-severe dysplasia than controls.

Furthermore, ROC curve analysis reveals salivary TNF-α to be a better marker for detecting OSCC. The sensitivity and specificity of TNF-α were found to be 93.3% each in the present study.

There is a difference in TNF-α value between studies which can be attributable to variation in the assay kit and techniques used. There is variation in the detection range of different ELISA kits. Cayman's TNF-α assay used in the study permits TNF-α measurements within the range of 0–250 pg/ml, typically with a limit of detection of 3.9 pg/ml. Hence, standardization of technique is recommended.

The results of the study show that the patients with leukoplakia and OSCC have increase in production of TNF-α which might indicate an altered immune response. TNF-α may be a potential monitoring molecule for the transformation of premalignancy to malignancy.[9],[13] The results of the present study also substantiate the observation of earlier reports. Oral cavity is subjected to various sources of inflammation such as dental plaque, trauma, and periodontitis. TNF-α, a proinflammatory cytokine, is elevated in chronic periodontitis,[18] and hence, in the present study, patients with chronic periodontitis were excluded which enhanced the validity of TNF-α as biomarker. It has been concluded that to monitor malignant transformation of leukoplakia further follow-up studies with larger sample is required. Salivary biomarkers, IL-1 β, IL-6, IL-8, MIP-1 β, eotaxin, and interferon-gamma and TNF-α showed significant differences between OSCC patients and controls in a study by Lee et al.,[19] and hence future studies, including these potential markers as a panel of biomarkers to aid in the early diagnosis of OSCC can be carried out.

Over the years, numerous biomarkers for OSCC have been suggested to foresee the prognosis of OSCC patients,[20] but there is still an ongoing debate regarding their clinical implementation.[21] Clinical trials could be obtained to prove clinical importance of TNF-α for survival rate, tumor recurrence, nodal metastasis, and resistance of therapy. Further clinical investigations are, therefore, required.


We would like to thank Colgate Palmolive India limited for their financial assistance in conducting the study.

Financial support and sponsorship


Conflicts of interest

There are no conflicts of interest.


1Khurshid Z, Zafar MS, Khan RS, Najeeb S, Slowey PD, Rehman IU. Role of salivary biomarkers in oral cancer detection. Adv Clin Chem 2018;86:23-70.
2Radhika T, Jeddy N, Nithya S, Muthumeenakshi RM. Salivary biomarkers in oral squamous cell carcinoma – An insight. J Oral Biol Craniofac Res 2016;6:S51-4.
3Krishna Prasad RB, Sharma A, Babu HM. An insight into salivary markers in oral cancer. Dent Res J (Isfahan) 2013;10:287-95.
4Landskron G, De la Fuente M, Thuwajit P, Thuwajit C, Hermoso MA. Chronic inflammation and cytokines in the tumor microenvironment. J Immunol Res 2014;2014:149185.
5Wang X, Lin Y. Tumor necrosis factor and cancer, buddies or foes? Acta Pharmacol Sin 2008;29:1275-88.
6Wiebe CB, Putnins EE. The periodontal disease classification system of the American Academy of Periodontology – an update. J Can Dent Assoc 2000;66:594-7.
7Cheng YS, Rees T, Wright J. A review of research on salivary biomarkers for oral cancer detection. Clin Transl Med 2014;3:3.
8Rhodus NL, Ho V, Miller CS, Myers S, Ondrey F. NF-kappaB dependent cytokine levels in saliva of patients with oral preneoplastic lesions and oral squamous cell carcinoma. Cancer Detect Prev 2005;29:42-5.
9Juretić M, Cerović R, Belušić-Gobić M, Brekalo Pršo I, Kqiku L, Špalj S, et al. Salivary levels of TNF-α and IL-6 in patients with oral premalignant and malignant lesions. Folia Biol (Praha) 2013;59:99-102.
10Brailo V, Vucićević-Boras V, Cekić-Arambasin A, Alajbeg IZ, Milenović A, Lukac J. The significance of salivary interleukin 6 and tumor necrosis factor alpha in patients with oral leukoplakia. Oral Oncol 2006;42:370-3.
11Krishnan R, Thayalan DK, Padmanaban R, Ramadas R, Annasamy RK, Anandan N. Association of serum and salivary tumor necrosis factor-α with histological grading in oral cancer and its role in differentiating premalignant and malignant oral disease. Asian Pac J Cancer Prev 2013;15:7141-8.
12SahebJamee M, Eslami M, AtarbashiMoghadam F, Sarafnejad A. Salivary concentration of TNFalpha, IL1 alpha, IL6, and IL8 in oral squamous cell carcinoma. Med Oral Patol Oral Cir Bucal 2008;13:E292-5.
13Brailo V, Vucicevic-Boras V, Lukac J, Biocina-Lukenda D, Zilic-Alajbeg I, Milenovic A, et al. Salivary and serum interleukin 1 beta, interleukin 6 and tumor necrosis factor alpha in patients with leukoplakia and oral cancer. Med Oral Patol Oral Cir Bucal 2012;17:e10-5.
14Patil S, Kaswan S, Rahman F, Dhoni B, Wadhawan R. Salivary and serum tumor necrosis factor alpha, interleukin 1 alpha, interleukin 1 beta, interleukin 6 and interleukin 8 in patients with oral carcinoma and leukoplakia. J Nepal Dent Assoc 2013;13:33-8.
15Kurokawa H, Yamashita M, Yamashita Y, Murata T, Miura K, Kajiyama M. Estimation of tumor necrosis factor-alpha in the diagnosis, the prognosis and the treatment follow-up of oral squamous cell carcinoma. Fukuoka Igaku Zasshi 1998;89:312-20.
16Korostoff A, Reder L, Masood R, Sinha UK. The role of salivary cytokine biomarkers in tongue cancer invasion and mortality. Oral Oncol 2011;47:282-7.
17Rhodus NL, Cheng B, Myers S, Miller L, Ho V, Ondrey F. The feasibility of monitoring NF-kappaB associated cytokines: TNF-alpha, IL-1alpha, IL-6, and IL-8 in whole saliva for the malignant transformation of oral lichen planus. Mol Carcinog 2005;44:77-82.
18Thilagar S, Theyagarajan R, Sudhakar U, Suresh S, Saketharaman P, Ahamed N. Comparison of serum tumor necrosis factor-α levels in rheumatoid arthritis individuals with and without chronic periodontitis: A biochemical study. J Indian Soc Periodontol 2018;22:116-21.
19Lee LT, Wong YK, Hsiao HY, Wang YW, Chan MY, Chang KW. Evaluation of saliva and plasma cytokine biomarkers in patients with oral squamous cell carcinoma. Int J Oral Maxillofac Surg 2018;47:699-707.
20Rivera C, Oliveira AK, Costa RAP, De Rossi T, Paes Leme AF. Prognostic biomarkers in oral squamous cell carcinoma: A systematic review. Oral Oncol 2017;72:38-47.
21Blatt S, Krüger M, Ziebart T, Sagheb K, Schiegnitz E, Goetze E, et al. Biomarkers in diagnosis and therapy of oral squamous cell carcinoma: A review of the literature. J Craniomaxillofac Surg 2017;45:722-30.