Estimation of Periostin and Tumour Necrosis Factor-α in Type II Diabetics with Chronic Periodontitis: A case-control study.

BACKGROUND: Periostin, a matricellular protein, is downregulated in chronic inflammatory periodontal disease and is negatively modulated by tumor necrosis factor-α (TNF-α) in human periodontal fibroblast cell culture. The study aimed to estimate the gingival crevicular fluid (GCF) levels of periostin and TNF-α and to discern their relationship in chronic periodontitis (CP) individuals with and without Type II diabetes mellitus (DM).
MATERIALS AND METHODS: A total of 60 participants were divided into three groups, with 20 in each group. Group I - systemically and periodontally healthy, Group II - generalized CP, and Group III - generalized CP with Type II DM. Plaque index, gingival index, sulcular bleeding index, probing depth, and clinical attachment level were recorded. GCF periostin and TNF-α were quantified using the enzyme-linked immunosorbent assay.
RESULTS: Intergroup comparison was performed using the one-way ANOVA and Kruskal-Wallis. The relationship between the variables was analyzed using the Pearson's and Kendall's Tau correlation. The GCF periostin levels in Groups I, II, and III was 27.52 ± 2.39 ng/mL, 20.18 ± 1.42 ng/mL, and 16.77 ± 3.29 ng/mL, respectively. The GCF TNF-α levels in Groups I, II, and III was 92.41 ± 19.30 ng/L, 118.53 ± 21.93 ng/L, and 147.67 ± 16.35 ng/L, respectively. Periostin decreased, and TNF-α increased in periodontal disease; moreover, periostin level correlated negatively with all the site-specific clinical parameters whereas TNF-α positively correlated (P < 0.001).
CONCLUSIONS: TNF-α strongly and negatively downregulates periostin in a chronically inflamed locale leading to compromised integrity of the periodontium.

INTRODUCTION

The extracellular matrix (ECM) is ubiquitously present within all the tissues and shows the evidence of bidirectional interaction with the cells that secrete it based on the concept of dynamic reciprocity by Bornstein et al. periostin, a 90 kDa matricellular protein with site-specific expression in the periosteum and periodontal ligament (PDL) was identified by Takeshita et al.[12]

Periostin is primarily secreted by the periodontal fibroblasts and has been immunolocalized on the cytoplasmic extensions of the periodontal fibroblast in areas closely associated with collagen fiber bundles. Thus, suggesting its role in cell-matrix interactions possibly resulting in remodeling of ECM and also as an adhesion molecule to withstand mechanical stresses, thereby contributing to strength and rigidity of the PDL.[3] Periostin null mice models showed severe incisor enamel defects, early-onset periodontal disease such as phenotype and dwarfism, aberrant collagen fibril maturation and assembly, as well as disorganized collagen cross-linking.[4] Therefore, it can be contemplated that in health periostin attempts to maintain the integrity of the healthy PDL.

Periodontitis is characterized by chronically inflamed microenvironment contributing to the degradation of ECM and alveolar bone.[5] The expression of periostin is shown to be downregulated in periodontal disease attributed either to the decrease in the number of fibroblast or diminished secretion of periostin by the fibroblasts in an inflamed milieu. On the other hand, in disease conditions elsewhere in the body, periostin takes up a totally different role and behaves analogous to a catabolic protein by modulating the secretion of proinflammatory cytokines.[6] Chronic exposure of human PDL cells to the tumor necrosis factor-α (TNF-α) and Porphyromonas gingivalis results in reduced periostin expression and incorporation into the ECM.[7] Therefore, it can be suggested that periostin and TNF-α possibly have a paradoxical role in the periodontium, with the former aiming to maintain a stable periodontium and the latter involved in the progression of periodontal disease.

Type II diabetes mellitus (DM) is a hyperinflammatory state associated with overt production of proinflammatory cytokines and is frequently associated with rapid and severe periodontal tissue destruction; similarly, periodontal inflammation worsens glycemic control, resulting in a two-way relationship.[89] Coexistence of both the diseases can further upregulate the inflammatory response. However, till date, no clinical studies have quantified periostin expression in Type II diabetics with chronic periodontal disease.

Based on the limited scientific evidence, there is a basic understanding that periostin is downregulated in the presence of chronic inflammatory periodontal disease and TNF-α modulates the expression of periostin.[6] However, there are inconsistencies which warrant further exploration. To the best of our knowledge, no clinical studies have investigated the relationship between periostin and TNF-α level in periodontal health and in disease states. Hence, in this study, we aimed to explore the relationship between these biological mediators. Considering the common inflammatory background and hyperinflammatory state prevailing in participants with coexisting Type II DM and periodontal disease, it would be interesting to investigate whether periostin levels would be further downregulated in the presence of proinflammatory cytokine-like TNF-α.

MATERIALS AND METHODS

Ethical clearance

Study participants for this case–control study were recruited from the outpatient clinic of the Department of Periodontics, from a dental educational institution. The study period was from January 2016 to July 2017. The approval was obtained from the Institutional Scientific and Ethical Committee Review Board (SRMDC/IRB/2015/MDS/MO.504).

Sample size calculation

The sample size was calculated based on the study by Balli et al. 2015.[10] Based on the power analysis with α error of 5% and 95% power, the preferred sample size for this study was 39. The ultimate sample size was increased to 60 with 20 participants in each group to have sufficiently acceptable statistical power and significance.

Subject allotment into groups

Convenience sampling was followed, and the participants were allotted into three groups based on the specific criteria as follows: Group I: systemically and periodontally healthy (PH), Group II: generalized chronic periodontitis (GCP), and Group III: generalized CP with type II DM (GCP-DM).

Group wise inclusion criteria for enrolment of subjects

Systemically and PH participants with probing pocket depth (PPD) ≤3 mm with no evidence of clinical attachment loss (CAL) and bleeding on probing with gingival index (GI) = 0 were assigned into Group I; participants with clinical signs of gingival inflammation with a GI ≥1 and PPD ≥5 mm with CAL >2 mm in more than 30% of sites with moderate-to-severe CP and at least 20 teeth remaining excluding third molars with 10 teeth in each jaw were allotted into Group II;[11] and participants with GCP (selected based on Group II criteria) diagnosed with Type II diabetes for a minimum of 2 years, under diet control and oral hypoglycemic medications with good or fair glycemic control (confirmed with hemoglobin A1C [HbA1C] <8%) were allotted into Group III.[12]

Participants were not recruited if they were under long-term anti-inflammatory medications or contraceptives or any drug that could affect periodontal status. Those who have undergone periodontal therapy previously or were affected with conditions requiring prophylactic antibiotic coverage for routine dental therapy were excluded from the study. In addition, smokers, chronic alcoholics, pregnant and lactating mothers, participants with any acute, chronic or allergic diseases, and those with poorly controlled DM (HbA1C >8%) with a history of diabetic complications were not included.

Clinical examination

The PI,[13] GI,[13] and sulcular bleeding index[14] were recorded; scoring was done on all teeth present, i.e., four sites per tooth (mesiobuccal, midbuccal, distobuccal, and palatal/lingual). PPD and CAL were assessed around six sites per tooth (mesiobuccal, midbuccal, distobuccal, mesiolingual/palatal, midlingual/palatal, and distolingual/palatal) using UNC-15 periodontal probe on the first visit.

Gingival crevicular fluid sample collection

Gingival crevicular fluid (GCF) was collected from the healthy sites with no clinical inflammation in the first appointment for Group I. Similarly, a site with deepest PPD was chosen for GCF collection from both Groups II and III. For Group III participants, in addition, 2 ml of blood was drawn for enzymatic HbA1C assay analysis in the first appointment followed by sample collection in the second appointment.

Participants were made to sit comfortably in the upright position on the dental chair with proper illumination, and the collection was done 2–3 h after breakfast. The sites were air dried and isolated with cotton rolls and suction to prevent contamination with saliva. Supragingival plaque and calculus if present were removed using a sterile curette. Calibrated, volumetric, microcapillary pipette (Sigma Aldrich, St Louis, MO, USA) was placed at the gingival margin, and 5 μl of GCF was collected. Once collected, the samples were immediately transferred to an Eppendorf tube and stored at –80°C in a deep freezer, until further processing. Blood contaminated samples were discarded.

Biochemical assay

Analysis of gingival crevicular fluid levels of Periostin and tumor necrosis factor-α by enzyme-linked immunosorbent assay

Periostin and TNF-α level in GCF were analyzed using commercially available enzyme-linked immunosorbent assay kit as per the manufacturer instruction (Bioassay Technology Laboratory, 1713, Junjiang International Bldg, Yangpu Dist, Shanghai, China). The microtiter plate provided with the kit was precoated with an antibody specific to periostin. Standards or samples were added to the appropriate microtiter plate wells, and antihuman periostin antibody was added to bind human periostin and incubated for 60 min at 37°C. The unbound biotin-conjugated antihuman periostin antibody was washed away, and Streptavidin-Horseradish peroxidase (HRP) was added and incubated. The Streptavidin-HRP binds to the biotin-conjugated antihuman periostin antibody following this; the unbound enzyme was washed away. The enzyme-substrate reaction was terminated by addition of sulfuric acid solution, and the color change was measured at a wavelength of 450 nm using microwell plate reader (Thermoscientific Multiscan Ex, 2010, USA). The periostin levels were determined by comparing the optic density of the samples to the standard curve. Sensitivity assay of the kit was 0.251 ng/ml. The mean interassay coefficient of variation (CV) % and intraassay CV % for periostin were <10% and <8%, respectively. The periostin concentration was expressed as ng/ml.

Similarly, TNF-α was estimated as per the manufacturer instruction. The sensitivity assay was 1.52 ng/L. The mean interassay CV % and intraassay CV % for TNF-α were <10% and <8%10, respectively. The TNF-α concentration was expressed as ng/L.

RESULTS

The study included 31 males (51.6%) and 29 females (48.3%), allotted into three groups. Groups I, II, and III included 10 males (50%) and 10 females (50%), 9 males (45%) and 11 females (55%), and 12 males (60%) and 8 females (40%), respectively. The mean age of the participants was 36.39 ± 13.77 years for males and 35.07 ± 12.10 years for females.

Table 1 shows the descriptive statistics of the full-mouth and site-specific clinical parameters, along with the evaluated biochemical parameters. Periostin levels were highest in the healthy group and least in Group III [Figure 1] suggestive of its downregulation in periodontally diseased sites. Contrarily, TNF-α levels were increased in Group III when compared with healthy, and Group II demonstrating the hyperinflammatory state persisting in those participants with coexisting generalized CP with Type II diabetes [Figure 2].

Table 1

Mean and standard deviation of the clinical and biochemical parameters evaluated

Variables	Mean±SD			P
	Group I (n=20)	Group II (n=20)	Group III (n=20)
Full-mouth GI#	0.008±0.01	1.67±0.41	1.66±0.54	<0.001**
Full-mouth PI#	0.13±0.20	1.37±0.30	1.26±0.43	<0.001**
Full-mouth SBI¶	0.00±0.00	2.00±0.57	1.75±0.64	<0.001**
Full-mouth PPD¶ (mm)	1.65±0.33	5.52±0.16	5.63±0.52	<0.001**
Full-mouth CAL¶ (mm)	0.00±0.00	2.88±0.81	2.80±0.75	<0.001**
Sampled site GI#	0.00±0.00	1.90±0.64	2.00±0.64	<0.001**
Sampled site PI#	0.40±0.30	1.75±0.55	1.65±0.58	<0.001**
Sampled site SBI#	0.00±0.00	2.55±0.82	2.35±0.67	<0.001**
Sampled site PPD# (mm)	1.45±0.60	8.90±2.33	10.05±2.83	<0.001**
Sampled site CAL# (mm)	0.00±0.00	6.65±3.32	7.40±3.20	<0.001**
Periostin¶ (ng/mL)	27.52±2.39	20.18±1.42	16.77±3.29	<0.001**
TNF-α¶ (ng/L)	92.41±19.30	118.53±21.93	147.67±16.35	<0.001**

Full-mouth SBI, PPD, CAL. ¶ANOVA (P≤0.001). Full-mouth PI, GI, sampled site-specific GI, PI, SBI, PPD, and CAL. #Mann–Whitney U-test (P=<0.001). **P<0.01 - Highly significant. GI – Gingival index; PI – Plaque index; SBI – Sulcular bleeding index; PPD – Probing pocket depth; CAL – Clinical attachment level; TNF – Tumor necrosis factor; SD – Standard deviation; n – number; P – Probability of occurrence

Figure 1

Mean periostin levels (ng/mL) in the three groups. GCP – Generalized chronic periodontitis; GCP + DM – Generalized chronic periodontitis with diabetes mellitus

Figure 2

Mean TNF-α levels (ng/L) in the three groups. GCP – Generalized chronic periodontitis; GCP + DM – Generalized chronic periodontitis with diabetes mellitus; TNF – Tumor necrosis factor

Intergroup comparison of the clinical parameters showed statistical significance P < 0.001 in sampled site-specific GI, PI, SBI, PPD, and CAL between Groups I and II (P < 0.001) and Groups I and III (P < 0.001). However, no significant difference was seen between the diseased groups (Groups II and III) [Table 2].

Table 2

Mann-Whitney tests for multiple pairwise comparisons of site-specific clinical parameters

Variable assessed	Group	P
Sampled site GI	Group I versus Group II	<0.001**
	Group I versus Group III	<0.001**
	Group II versus Group III	0.62
Sampled site PI	Group I versus Group II	<0.001**
	Group I versus Group III	<0.001**
	Group II versus Group III	0.55
Sampled site SBI	Group I versus Group II	<0.001**
	Group I versus Group III	<0.001**
	Group II versus Group III	0.33
Sampled site PPD (mm)	Group I versus Group II	<0.001**
	Group I versus Group III	<0.001**
	Group II versus Group III	0.19
Sampled site CAL (mm)	Group I versus Group II	<0.001**
	Group I versus Group III	<0.001**
	Group II versus Group III	0.38

**P<0.01 - highly significant. GI – Gingival Index; PI – Plaque Index; SBI – Sulcular Bleeding Index; PPD – Probing Pocket Depth; CAL – Clinical Attachment Level; P – Probability of occurrence

Table 3 represents the pairwise comparison between groups for periostin and TNF-α, showing statistical significance (P < 0.001) between the three groups. Negative correlation was observed between the clinical parameters – GI, PI, SBI, PPD and CAL, and periostin levels. Similarly, a positive correlation was seen between clinical parameters and TNF-α level [Table 4]. Periostin and TNF-α showed strong negative correlation with statistical significance (P < 0.001) [Figure 3].

Table 3

Tukey’s post hoc following the one-way ANOVA analysis for pairwise comparison between groups for periostin and tumor necrosis factor-alpha levels

Variable	Pair-wise comparison	Mean difference	P
Periostin (ng/mL)	Group I versus Group II	7.33	<0.001**
	Group I versus Group III	10.74	<0.001**
	Group II versus Group III	3.41	<0.001**
TNF-α (ng/L)	Group I versus Group II	−26.1	<0.001**
	Group I versus Group III	−55.2	<0.001**
	Group II versus Group III	−29.1	<0.001**

**P<0.01 - Highly significant. TNF – Tumor necrosis factor; P – Probability of occurrence

Table 4

Kendall’s-tau correlation of site-specific clinical parameters assessed with periostin and tumor necrosis factor-α levels

	Group I		Group II		Group III		Total
	r	P	r	P	r	P	r	P
Sampled site GI versus Periostin	NA	NA	0.03	>0.05	0.33	>0.05	−0.48	<0.001**
Sampled site GI versus TNF-α	NA	NA	−0.06	>0.05	−0.15	>0.05	0.44	<0.001**
Sampled site PI versus Periostin	0.10	>0.05	0.07	>0.05	0.26	>0.05	−0.40	<0.001**
Sampled site PI versus TNF-α	−0.16	>0.05	−0.07	>0.05	−0.04	>0.05	0.34	<0.01**
Sampled site SBI versus Periostin	NA	NA	0.01	>0.05	0.21	>0.05	−0.45	<0.001**
Sampled site SBI versus TNF-α	NA	NA	−0.10	>0.05	0.02	>0.05	0.39	<0.001**
Sampled site PPD (mm) versus Periostin	0.10	>0.05	−0.02	>0.05	0.05	>0.05	−0.52	<0.001**
Sampled site PPD (mm) versus TNF-α	−0.16	>0.05	−0.07	>0.05	0.01	>0.05	0.42	<0.001**
Sampled site CAL (mm) versus Periostin	NA	NA	−0.11	>0.05	0.11	>0.05	−0.53	<0.001**
Sampled site CAL (mm) versus TNF-α	NA	NA	−0.19	>0.05	−0.04	>0.05	0.40	<0.001**
Periostin versus TNF-α	−0.1	>0.05	0.12	>0.05	0.17	>0.05	−0.46	<0.001**

**P<0.01 - Highly significant. GI – Gingival Index; PI – Plaque index; SBI – Sulcular bleeding index; PPD – Probing pocket depth; CAL – Clinical attachment Level; TNF – Tumor necrosis factor; NA – Not available; r – Correlation coefficient; P – Probability of occurrence

Figure 3

Correlation between periostin and TNF-α levels. TNF – Tumor necrosis factor

Table 5 depicts the fitted regression model for periostin, based on the adjusted R2. The independent variables, site-specific GI, and PD accounted for 45% and 56% variation of periostin levels, respectively. Similarly, SBI, CAL, and PI accounted for 46%, 43%, and 29% variation of periostin levels. Similarly, the fitted regression model for TNF-α, based on the adjusted R2, the independent variables, site-specific GI, PD, CAL, SBI, and PI account for 28%, 33%, 21%, 27%, and 14% variation of TNF-α levels, respectively.

Table 5

Simple linear regression model to predict periostin and tumor necrosis factor-α levels based on site-specific clinical parameters

Variable	Fitted regression model	Adjusted R2	P
Sampled site GI	Periostin=25.78-3.30 × GI site	0.45	<0.001**
Sampled site PD	Periostin=27.49-0.88 × PD site	0.56	<0.001**
Sampled site CAL	Periostin=25.27-0.80 × CAL site	0.43	<0.001**
Sampled site SBI	Periostin=25.89-2.69 × SBI site	0.46	<0.001**
Sampled site PI	Periostin=25.90-3.48 × PI site	0.29	<0.001**
Sampled site GI	TNF-α=99.64 +15.30 × GI	0.28	<0.001**
Sampled site PD (mm)	TNF-α=92.38 + 3.99 × PD	0.33	<0.001**
Sampled site CAL (mm)	TNF-α=103.93 + 3.33 × CAL	0.21	<0.001**
Sampled site SBI	TNF-α=99.65+12.17 × SBI	0.27	<0.001**
Sampled site PI	TNF-α=101.07+14.58 × PI	0.14	<0.001**

**P<0.01 - Highly significant. GI – Gingival Index; PI – Plaque index; SBI – Sulcular bleeding index; PPD – Probing pocket depth; CAL – Clinical attachment Level; TNF – Tumor necrosis factor; R2 – Coefficient of determination; P – Probability of occurrence

DISCUSSION

In vitro studies on periodontal cell cultures have shown an inverse relationship between periostin and TNF-α.[7] Considering the paradoxical actions of both TNF-α and periostin in the periodontal disease, it would be interesting to use these two proteins as a “dual-biomarker” for assessing the periodontal disease activity.

The present study results showed that GCF periostin levels were the highest in Group I, lower in Group II, and least in Group III. There was a significant difference in the periostin levels between the three groups (P < 0.001) indicating its role in tissue homeostatic mechanisms for maintaining the periodontal health and its downregulation in the presence of inflammation. Periostin is a secreted ECM protein that is found in areas of normal fibrogenesis or pathologic fibrosis. In the periodontium, it is secreted predominantly by the gingival and periodontal fibroblasts contributing to the formation of a stable and mature PDL matrix. Hence, the predominance of periostin in the healthy sites is normal as it maintains the homeostasis in these tissues. However, in periodontally diseased sites, the levels reduce possibly attributed to the decreased number of fibroblasts or due to reduced secretion of periostin in the presence of overexpressed proinflammatory cytokines.[1516]

Balli et al. found a similar trend wherein the expression of periostin decreased significantly from health to disease (P < 0.05).[10] Periostin levels reduced strongly and negatively in an inflamed periodontium, further signifying its role in maintaining periodontal tissue integrity. Further, a negative correlation was observed between periostin and clinical parameters such as GI and CAL on considering the overall study participants ignoring the grouping; these findings were in line with our results (P < 0.001). Thus, it can be inferred that with increasing severity of periodontal disease and gingival inflammation, periostin levels decrease.

The findings were also in agreement with Aral et al., further validating the negative modulation of periostin expression by the chronically inflamed periodontium.[17] Kumaresan et al. estimated GCF periostin levels following nonsurgical periodontal therapy with adjunctive low-level laser therapy and observed that the levels improved following therapy demonstrating the role of inflammation.[18] Therefore, the authors concluded that periostin could be used as a biomarker to detect the disease activity, predict the disease progression, and evaluate the response following periodontal therapy.[18]

Padial-Molina et al. in their quest to identify the role of periostin during periodontal wound healing following open flap debridement carried out a surgical intervention study and observed varied expression of GCF periostin characterized by transient increase at 2 weeks following surgical therapy, and after 4 weeks, the values reverted back to baseline levels. Therefore, the authors hypothesized that the transient increase in the periostin levels assisted in the healing process, possibly due to its cell mitogenic property. The reduction in periostin level during wound maturation could be associated with the high rate of deposition of periostin in the ECM as the collagen structure matures. Therefore, in the context of wound healing, their observation propounded that the expression of periostin might have key implications following surgery in the restoration of the original architecture and function of the periodontium, implicating its prospective role as a regenerative agent.[19]

On the other contrary, Akman et al. observed reduced GCF periostin levels in PH sites when compared with inflamed sites around the natural tooth.[20] This was in contradiction to all the previous studies wherein the healthy sites showed higher expression of periostin.[101718] Nevertheless, keeping in mind the unique milieu of peri-implant area, we theorize that the different microenvironment characterized by parallel arrangement of collagen fibers to the implant surface, the reduced number of fibroblasts and reduction in vascularity could have potentially affected the expression of periostin accounting for such a varied expression. Furthermore, periostin is primarily produced by fibroblasts; a reduction in the fibroblast content in the peri-implant space could have favored the production of periostin from the osteoblasts in the periosteum and from the existing gingival fibroblasts.

In this study, GCP patients with Type II DM showed the least periostin expression. Currently, no other scientific literature is available to corroborate our findings; however, keeping in mind the hyperinflammatory state existing in both the diseases, it can be suggested that an increased inflammatory burden could have resulted in lower periostin expression in this group. Scientific literature associating systemic periostin levels with diabetes and its complications conversely demonstrate increased levels.[2122] Luo et al. showed elevated plasma periostin levels among diabetics, and the highest levels were observed among participants with both obesity and diabetes and the levels positively correlated with the inflammatory cytokines TNF-α and IL-6.[23] The disparity in the periostin expression between systemic and periodontal disease may be ascribed to the completely different environmental condition persisting in these diseases. Moreover, the gingival fibroblasts are phenotypically distinct and sustain inflammation, thereby affecting periostin expression.[24]

GCF TNF-α level varied significantly between the individual groups. The clinical parameters correlated positively and significantly with TNF-α level when the data from all the participants (60 in total) were grouped together indicating that with increasing severity of periodontal disease, TNF-α increased. The healthy patients (Group I) had lower levels of TNF-α when compared with GCP group. Furthermore, the highest expression was seen in GCP with Type II DM.

To the best of our knowledge, no clinical studies have evaluated the relationship between GCF periostin and TNF-α in CP and Type II DM until date. However, Padial-Molina et al., in their in vitro study, demonstrated the effect of pathogen-related virulence factors such as lipopolysaccharide (LPS) and proinflammatory cytokine TNF-α on the expression of periostin in PDL cells and explained that chronic exposure to proinflammatory cytokines or microbial virulence factors significantly reduced periostin levels in the loaded cultures. Periostin had a positive effect on cell proliferation, migration, and activation of survival signaling pathways, however, in the presence of TNF-α and bacterial LPS; a negative effect was observed on the above-said cellular processes. The specific mechanism of periostin reduction by bacterial by-products and/or the inflammatory mediators could not be explained.[25]

The periostin negatively correlated with TNF-α in the present study. Further, this inverse relationship was seen both within the groups and as well as when all the total study participants (60) were grouped together (P < 0.001). Thus, it can be inferred that in the absence of inflammation or reduced inflammation, elevated levels of periostin was associated with reduced levels of TNF-α, and in the presence of inflammation, the reverse was true. Periostin is a versatile protein, capable of functioning as anabolic or a catabolic molecule. It amplifies the inflammatory burden in systemic diseases by secreting proinflammatory cytokines and by initiating fibrosis.[26] However, in periodontal disease, the expression of periostin is downregulated in the presence of proinflammatory cytokines such as TNF-α. The mechanism for such a downregulation by TNF-α is yet unexplored. Based on their relationship in health and disease, simultaneous estimation of TNF-α and periostin in periodontal disease could aid in assessing the severity of periodontal disease.

Simple linear regression to predict the periostin and TNF-α level based on sampled sites GI, PI, SBI, PPD, and CAL of all the study participants (60) revealed that each of these clinical parameters predicted the variation in periostin and TNF-α levels independently and significantly. Sampled site PPD, GI, SBI, and CAL accounted for larger variation in periostin levels whereas TNF-α level was weakly predicted by these parameters. In addition, multiple regression revealed that among all the clinical variables assessed, only sampled site PPD accounted for their variation, possibly indicating that the expressions of periostin and TNF-α varied with the periodontal disease severity.

CONCLUSIONS

The results were in agreement with the study hypothesis wherein GCF periostin expression was reduced in GCP participants, with and without Type II DM. TNF-α was elevated in the diseased groups and the levels inversely correlated with periostin. Thus, periostin and TNF-α play a significant role in the periodontal disease activity, and the negative modulation of periostin by TNF-α indicates their reciprocal and collective action on the progression of periodontal disease, and in the future, possibly could contribute toward the development of “dual-markers” to appraise the severity and progression of periodontal destruction. However, prospective studies on a larger and diverse population are required to further validate these findings, so that these biomarkers can be applied in the clinical scenario.

Financial support and sponsorship

This was a self-funded study

Conflicts of interest

There are no conflicts of interest.