Salivary TNF-alpha: A potential marker of periodontal destruction

Abstract

Aims and Objectives:

(1) To evaluate the effect of type 2 diabetes mellitus on salivary TNF-α level in chronic periodontitis. (2) To evaluate the effect of smoking on salivary TNF-α level in chronic periodontitis. (3) To compare and correlate TNF-α level with the healthy individuals.

Materials and Methods:

Subjects aged 30-35 years were included for the study and divided into four groups as a group of 20 systemically and periodontally healthy individuals (group I), a group of 20 subjects with pocket probing depth (PPD) ≥5 mm and clinical attachment loss (CAL) of ≥2 mm (group II), a group of 20 diabetic subjects (of more than 5 years) with periodontal parameters as of group II as (group III) and a group of 20 subjects smoking (≥10 cigarettes a day) with periodontal parameters of group II as (group IV). Periodontal parameters of PPD, CAL, gingival index (GI), and plaque index (PI) were measured using standard indices and criteria. Three milliliter of unstimulated saliva was taken and salivary TNF-α determined by using ELISA technique (Quantikine Human total TNF-A immunoassay kit).

Results:

Data revealed highest mean TNF-α in group III followed by group IV, group II, and group I. Mean TNF-α of both group III (76.1%) and group IV (48.8%) was significantly higher as compared to group I (P < 0.001). Mean TNF-α of group III was also found to be significantly different and higher (68.1%) as compared to group II (P < 0.001). Although higher mean TNF-α (31.5%) was found in group IV in comparison to group II, the difference was not statistically significant. Besides above, TNF-α also showed a direct positive correlation with PPD in group II (r = 0.30, P > 0.05) and a significant negative correlation was observed between CAL and TNF-α in group IV.

Conclusion:

Our study clearly underlines a profound impact of diabetes and smoking on salivary TNF-α in chronic periodontitis subjects in comparison to healthy subjects. Moreover, diabetes status increased TNF-α significantly in comparison to smoking in chronic periodontitis patients.

INTRODUCTION

Oral health is indispensable to overall healthy being. Man has been suffering from ailments of oral cavity since time immemorial. Oral diseases especially caries and periodontitis are known for their high prevalence and rapid morbidity. Periodontal diseases are a group of chronic, progressive bacterial infections resulting in inflammation and destruction of tooth supporting tissues.[1] The periodontal disease is known to have a complex pathogenesis with both bacterial and host factors contributing to the destruction of periodontium. Role of host immune response is most important factor in periodontitis as it determines both disease progression and severity.[2] Diabetes mellitus is one such widely prevalent endocrine disorder which is known to alter this delicate balance thereby causing exaggerated tissue destruction when challenged with chronic diseases like periodontitis.[3] In smokers too, a suppressed immune response favor increased periodontal destruction.[4,5] Hence, immune function in periodontitis is a specialized function where every single molecule can have ramifications on the resultant tissue protective and destructive response.

Difficulty in determining active disease and ongoing destruction in periodontal tissue by traditional diagnostic aids like probing depth and attachment loss has proved them to be inadequate in modern era of periodontal therapeutics.[6] Search for a biomarker for periodontitis has resulted in researchers trying out and finding new molecules that can guide a clinician in many a decision regarding the patient's condition.

Tumor necrosis factor-alpha (TNF-α) is a proinflammatory cytokine released by macrophages is known for its substantial role in periodontitis mediated bone loss.[7] This can be detected in saliva and gingival crevicular fluid (GCF) in both health and periodontitis.[8] Increased concentration observed in periodontitis correlate closely with the tissue destruction and immune response.[9] TNF-α also inhibits insulin transduction and contributes to insulin resistance in diabetes mellitus.[10] Enhanced expressions of serum TNF-α have been observed in smokers in rheumatoid arthritis and chronic obstructive pulmonary disease (COPD). Also, up-regulation of its expression in keratinocytes in chronic inflammatory skin diseases like psoriasis has also been observed.[11] This clearly indicates the role of TNF-α in smokers and chronic inflammation. Also, salivary cytokines are increasingly being correlated with periodontal status and oral inflammatory burden in recent times.[9,12] As systemic inflammation also influence salivary inflammatory burden, there is a need to quantify this influence to make salivary biomarker assessment meaningful. Thus, the purpose of our study is to evaluate the impact of smoking and type II diabetes mellitus on salivary TNF-α levels in chronic periodontitis patients.

MATERIALS AND METHODS

A total of 80 age and sex matched individuals in the age range of 30-35 years and having at least 20 teeth were taken into the study from the out-patient wing of department of periodontics. Complete intra oral and extra oral examination was performed and systemically and periodontally healthy individuals were taken in group I. Subjects with pocket probing depth (PPD) ≥5 mm and clinical attachment level (CAL) ≥2 mm were included in group II. Subjects with diabetes mellitus diagnosed for more than 5 years and current metabolic control above the normal range (glycated hemoglobin [HbA1c] test >6%, measure on visit) along with same periodontal parameters as in group II were taken into group III. Current smoker subjects currently smoking ≥10 cigarettes a day along with same periodontal parameters as of group II were included in group IV. Institution review board approved consent was obtained from all the subjects after explaining about the study.

Exclusion criteria

Subjects with acute and aggressive periodontal conditions like aggressive periodontitis and acute necrotizing ulcerative gingivitis (ANUG) were excluded from the study. Also, subjects having any complicating/confounding condition for example hematological, hormonal, allegric or other chronic conditions and subjects taking medication from past 3 months excluding diabetic patients on medications were excluded from the study. Pregnant and lactating subjects were not included in the study. Similarly, subjects who had undergone surgery in past 6 months were excluded from the study. Patients with oro-mucosal abnormality were excluded from study.

Periodontal assessment

A detailed history of each subject pertaining to age, sex, smoking status, systemic and dental disease was recorded. An assessment of PPD, CAL, gingival index (GI; Loe and Silness), plaque index (PI; Silness and Loe) was done using University of North Carolina-15 (UNC-15 probes Hu-friedy's USA) and by same examiner to avoid subjective bias.

Unstimulated saliva was taken from all the selected subjects. Subjects were told not to eat or drink for 2 h prior to the examination. Three milliliter of unstimulated whole expectorated saliva was collected from each subject into sterile 5 ml saliva collecting tubes according to the method described by Navazesh.[13] Then the sample was stored at −80°C. Salivary TNF-α levels were determined in duplicate for each subject using the Quantikine Human total TNF-α immunoassay kit employing an ELISA technique, provided by RandD systems, Minneapolis, USA.

Statistical analysis

Expecting at least 5.0% impact with 5.0% margin of error and 80.0% power with 1:1 ratio. Minimum sample size required was 18 per group (73 in total). Therefore, a sample size of 20 subjects was selected for each group. Data were summarized as mean ± SD. Groups were compared by one way analysis of variance (ANOVA) and the significance of mean difference between the groups was done by Tukey's post hoc test. Discrete groups were compared by Chi-square (χ²) test. Pearson correlation analysis was done to assess association between TNF-α and periodontal status. Simple linear regression analysis was done to assess relative association between TNF-α and periodontal status considering TNF-α the independent variable and periodontal status the dependent variable. A two-sided (α = 2) P < 0.05 was considered statistically significant. All analyses were performed on STATISTICA software (Windows version 6.0).

RESULTS

Analysis

Comparing the mean TNF-α level of four groups [Table 2], ANOVA revealed significantly different TNF-α level among the groups (F = 52.84, P < 0.001) Tukey test revealed that the mean TNF-α level of both group 3 (26.52 ± 8.52 vs. 6.33 ± 2.15, q = 16.16; P < 0.001) and group 4 (12.35 ± 5.16 vs. 6.33 ± 2.15, q = 4.82; P < 0.01) were significantly different and higher as compared to group 1. Further, the mean TNF-α level of group 3 was also found to be significantly different and higher as compared to both group 2 (26.52 ± 8.52 vs. 8.46 ± 4.60, q = 14.45; P < 0.001) and group 4 (26.52 ± 8.52 vs. 12.35 ± 5.16, q = 11.33; P < 0.001). However, the mean TNF-α level did not differed significantly (P > 0.05) between group 1 and group 2 (6.33 ± 2.15 vs. 8.46 ± 4.60, q = 1.71; P = 0.624) and group 2 and group 4 (8.46 ± 4.60 vs. 12.35 ± 5.16, q = 3.12; P = 0.132), i.e. found to be statistically the same. However, the mean TNF-α level of group 2, group 3 and group 4 were 25.2, 76.1, and 48.8%, higher respectively, as compared to group 1.

Table 2.

Comparison of study groups with controls: TNF-α level of four groups

Table 1.

Mean values of parameters in subject groups

Comparing the mean TNF-α level of three groups, ANOVA revealed significantly different TNF-α level among the groups (F = 45.02, P < 0.001). Further, Tukey test [Table 3] revealed that the mean TNF-α level of especially group 3 (26.52 ± 8.52 vs. 8.46 ± 4.60, q = 12.75; P < 0.001) was significantly different and higher (68.1%) as compared to group 2. Further, the mean TNF-α level of group 3 was also found to be significantly different and higher (53.4%) as compared to group 4 (26.52 ± 8.52 vs. 12.35 ± 5.16, q = 10.00; P < 0.001). However, the mean TNF-α level did not differed significantly (P > 0.05) between group 2 and group 4 (8.46 ± 4.60 vs. 12.35 ± 5.16, q = 2.75; P = 0.136) though it was 31.5% higher in group 4 as compared to group 2.

Table 3.

Significance (P value) of mean difference of TNF-α level between the groups by Tukey test

Correlation

In group 1, TNF-α did not showed significant correlation with periodontal status besides high negative (inverse) correlation especially with GI (r = −0.26, P > 0.05) and PPD (r = −0.29, P > 0.05). Similarly, in group 2, TNF-α also not showed significant correlation with periodontal status even there was a high positive (direct) correlation with PPD (r = 0.30, P > 0.05) while high negative correlation with CAL (r = −0.39, P > 0.05). Further, TNF-α did not correlate well in any of the periodontal status of group 3. In contrast, in group 4, TNF-α showed significant and negative correlation with CAL (r = −0.46, P < 0.05).

However, TNF-α of all subjects (group 1 + group 2 + group 3 + group 4) showed positive and significant correlation with their periodontal status; PI (r = 0.42, P < 0.001), GI (r = 0.44, P < 0.001), PPD (r = 0.44, P < 0.001) and CAL (r = 0.39, P < 0.001). Further, regression analysis revealed that the TNF-α alone can accounts 18.0, 19.0, 19.0, and 15.0% total variations of PI, GI, PPD and CAL, respectively.

DISCUSSION

The present study was carried out to evaluate the effect of type 2 diabetes mellitus and smoking on levels of salivary TNF-α in chronic periodontitis patients and to correlate them with healthy individuals. Periodontitis is a disease characterized by loss of connective tissue attachment and bone around the teeth in conjunction with the formation of periodontal pockets due to the apical migration of junctional epithelium.

TNF-α, also known as cachectin and TNFSF1A, is the prototypic ligand of the TNF superfamily. It is a pleiotropic molecule that plays a central role in inflammation, immune system development, apoptosis, and lipid metabolism. TNF-α is a proinflammatory cytokine released by macrophages is known for its substantial role in periodontitis mediated bone loss. In diabetes advanced glycation end-products (AGEs) are formed which are responsible for diabetic collagen cross-links, which could bind to macrophage receptors induce a cycle of cytokine (IL-1 and TNF-α) up regulation. These can, in turn, induce an elevated expression of matrix metalloproteinases (MMP) in diabetic periodontal tissues.[14] Smoking has been found to exert a major effect on the protective elements of immune response, resulting in the extent and severity of periodontal destruction.[15] Microbial flora may or may not be affected by smoking. However, the host response which is a more important link in the chain may be substantially affected. Smoking impairs the normal host response in bacterial clearance and neutralizing of infection. The granulocyte function in smokers with persistent periodontitis is impaired, these cells react to a bacterial challenge by releasing the serine proteases elastase and matrix metalloproteinase which are related to which are related to degradation of connective tissue.[16] Therefore, the present study was undertaken to examine the effect of type 2 diabetes mellitus and smoking on levels of TNF-α in chronic periodontitis patients.

Our study was conducted on unstimulated saliva. Majority of investigations have involved GCF, in biomarker research because of its close proximity to the site of release of molecular markers, making it site specific but at the same time, limiting its role to express status of other inactive sites. However, sampling of saliva solved this problem because it represents pooled concentrations obtained from all the sites in the mouth giving overall assessment of disease status and severity. Secondarily, GCF samples collection involves collection of miniscule amount of fluid on filter paper strips which require more sampling time. In addition blood, saliva and plaque products contamination persists as a problem.[17] These problems are practically non-existent in saliva where collection is easy, non-invasive, and rapid. Its measurement requires no special equipment and expertise.

In this study, subjects with any acute and chronic systemic conditions, such as bronchiectasis, asthma, and atherosclerosis inflammatory bowel disease were excluded as these conditions lead to rise of TNF-α on their own, which can lead to a confounding effect in the study.[18] Quantitative determination of salivary TNF-α in the present study was done by the double antibody sandwich ELISA method which is very sensitive method for detecting TNF-α in saliva. The minimum detectable dose of TNF-α ranged from 0.01-0.06 ng/ml.

In the present study, the mean gingival index was lowest for group IV (1.62 ± 0.11) which is consistent with the study of Feldman et al.(1983) and Perber et al. (1985) who showed that smokers with periodontal disease had less gingival bleeding when compared with non-smokers.[19,20] This may be explained by the fact that one of numerous tobacco smoke by-products and nicotine, exert local vasoconstriction, reducing blood flow, edema and act to inhibit what are normally early signs of periodontal problems by decreasing gingival inflammation, redness, and bleeding. Grossi et al. (1995) found that tobacco smoking was strongly associated with both attachment loss and bone loss.[21] It has also been suggested that reduced bleeding reflects an underlying disruption of the immune response which may account for the increased loss of clinical attachment and alveolar bone. These findings suggest a reduced reliance on the use of gingival bleeding as an indicator of gingival inflammation when assessing a smoker's periodontal health.

Our results showed that there was an increase in gingival index in subjects with type 2 diabetes mellitus as compared to controls (2.38 ± 0.26 vs. 0.54 ± 0.11) which is in accordance with the study done by Ervasti et al. (1985) where poorly controlled diabetics had mean gingival score of 77.9 ± 19.2 as compared to 61.9 ± 25.1 in control group, which was statistically significant (P < 0.01).[22] This may be due to the metabolic imbalances in the tissue which can lower the resistance of diabetics to infection and the influence of initiation, development and progression of periodontal disease. Impaired neutophil chemotaxis has been found in diabetic patients and may be another factor in decreased response to inflammation.

The mean probing depth of group IV (4.81 ± 0.42) was highest, which was in accordance with the study be Feldman et al. (1983) who reported that it may be the poor oral hygiene practices that are significantly related to the development of periodontal pockets in cigarette smoker.[19] Soder et al. (2002) also concluded that in periodontitis sites of smokers, the local host response to the bacterial challenge resulted in neutrophils degranulation, which is considered to be one of the major pathogenic mechanisms in periodontitis.[17] Due to the enhancement of degranulation in neutrophils, the smokers are more sensitive to bacterial challenge, with an increased risk of releasing elastase and of loss of attachment compared to non-smokers.

The present study showed that the mean clinical attachment level of group III (3.51 ± 0.39) to be the highest among all four groups. This may be explained by the fact that the increased levels of periodontal attachment and bone loss seen in diabetic patients may be associated with the alterations in connective tissue metabolism that uncouple the resorptive and formative responses. Impaired osseous healing and bone turnover in association with hyperglycemia have been demonstrated in a number of studies. The effects of a hyperglycemic state include inhibition of osteoblastic cell proliferation and collagen production that result in reduced bone formation and diminished mechanical properties of the newly formed bone. Lawrence J Emrich concluded that type 2 diabetes mellitus has an increased risk of destructive periodontitis.[23] Using attachment loss or bone loss as measure of destructive periodontitis, diabetic subjects showed a risk of 2.81 or 3.43 times higher than the non-diabetics. The odd ration for diabetic subjects was 3.43 (95% confidence interval 2.28 to 5.16) where bone loss was used to measure periodontal destruction. These findings demonstrate that diabetes increases the risk of developing destructive periodontal disease bout three-fold.

Comparing rise of TNF-α in group 1 and group 2 (6.33 ± 2.15 vs. 8.46 ± 4.60) which shows a higher mean salivary TNF-α in chronic periodontitis subjects as compared to controls is highly in conformity with the results obtained by Frodge (1998) where mean salivary TNF-α was found to be significantly higher in individuals with periodontal disease (4.33 vs. 2.03 pg/ml).[18] These results are also in agreement with the results obtained by Passoja et al. (2010) which showed elevated levels of TNF-α in subjects of chronic periodontitis albeit in serum as compared to controls.[24] This can be understood by the role of TNF-α as proinflammatory cytokine which regulate transcription of MMP's’ which manifest in tissue destruction of chronic periodontitis as stated by Birkedal-Hansen (1993). Our study showed that subjects with group 3 had significantly higher TNF-α (26.52 ± 8.52 vs. 6.33 ± 2.15, q = 16.16; P < 0.001) in comparison to group 1 which was 76.1% higher. Even after comparing group 3 with group 2 (26.52 ± 8.52 vs. 8.46 ± 4.60, q = 12.75; P < 0.001) a significant increase was noticed (68.1%). This clearly proves that diabetes alone contributed much to the difference found in salivary TNF-α as compared to the chronic periodontitis as seen when comparing group 1 with group 2 (6.33 ± 2.15 vs. 8.46 ± 4.60, q = 1.71; P = 0.624) which depicts a rise of 25.2%. Here a clear rise of 68.1% which can be attributed to diabetes alone comes in conformity of the results obtained by Yoon et al. (2012) where analysis resulted that both diabetes and periodontitis independently correlated positively with inflammatory markers in unstimulated saliva.[25] This result also comes in agreement with the results of Salvi, Beck and Offenbacher (1998) which found a rise in TNF-α (4.6-fold) as compared to non-diabetics controls in GCF. This increase was found to be independent of chronic periodontitis status. This proved that diabetes as significant contributor of proinflammatory cytokines like TNF-α as identified by Emrich, Shlossman, Genco (1991) in their results diabetes as an important parameter in periodontal destruction after adjusting for the effects of demographic variables and several indices of oral health.

Mean TNF-α was also significantly elevated in group 4 as compared to group 1 (12.35 ± 5.16 vs. 6.33 ± 2.15) which shows a rise of 48.8%. Even after comparing group 2 with group 4 (8.46 ± 4.60 vs. 12.35 ± 5.16) which shows a rise of 31.5%. This clearly proves that smoking too contributed to the difference found in slivary TNF-α as compared to the chronic periodontitis as seen when comparing group 1 with group 2 (6.33 ± 2.15 vs. 8.46 ± 4.60, q = 1.71; P = 0.624) which depicts a rise of 25.2%. Hence, a clear rise of 31.5% can be attributed to smoking which is in agreement with the results of Boström, Linder and Bergström (1998) in GCF where smokers with chronic periodontitis had an increased TNF-α in comparison to non-smokers. It was also noticed by Fredriksson, Bergström, and Åsman (2002) that TNF-α release from peripheral neutrophils in plasma seem to be influenced more by cigarette smoking than periodontal disease.[26]

Moreover, TNF-α of all subjects (group 1 + group 2 + group 3 + group 4) subjects showed positive and significant correlation with their periodontal status; PI (r = 0.42, P < 0.001), GI (r = 0.44, P < 0.001), PPD (r = 0.44, P < 0.001) and CAL (r = 0.39, P < 0.001). Further, regression analysis revealed that the TNF-α alone can accounts 18.0, 19.0, 19.0, and 15.0% total variations of PI, GI, PPD, and CAL, respectively. Although the results obtained clearly implicated diabetes and smoking to be major contributor to salivary TNF-α, long term longitudinal studies in all the four groups are required for important generalizations regarding role of TNF-α as valuable diagnostic biomarker. Also, this study fails to co-relate TNF-α with duration of smoking in years. Measurement of GCF TNF-α simultaneously could have provided another opportunity in quantifying the extent of periodontal contribution in salivary TNF-α burden and thereby exact quantification of systemic contribution by type 2 diabetes mellitus and smoking could have been obtained. In its limited extent, this study finds its value in actively establish a role of TNF-α as a diagnostic marker but also as a marker of disease risk and susceptibility. Moreover, it might even guide determine unnoticed systemic influence/condition masked under the veil of periodontitis. Therefore, using TNF-α level in saliva as a non-invasive biomarker holds strong potential especially that the dental field is moving toward personalized medicine.

CONCLUSION

It can be clearly observed that increased levels of salivary TNF-α were present in all the three diseased (test) groups. Salivary TNF-α of group III (type 2 diabetes Mellitus with chronic periodontitis) was significantly higher when compared to group I (control) and group II (chronic periodontitis) suggesting the association between diabetes and increased proinflammatory cytokine TNF-α. Also, salivary TNF-α of group IV (smoker with chronic periodontitis) was significantly higher when compared to group I (control) and group II (chronic periodontitis) suggesting that smoking interferes with the normal host defense mechanisms and stimulates destructive effects within the host. Hence, our study clearly underlines a profound impact of diabetes and smoking on salivary TNF-α in chronic periodontitis subjects in comparison to healthy subjects. Moreover, diabetes status increased TNF-α significantly in comparison to smoking in chronic periodontitis patients.