Association of IL-10 and TNF-α Polymorphisms with Dental Peri-Implant Disease Risk: A Meta-Analysis, Meta-Regression, and Trial Sequential Analysis

Genetic susceptibility has been reported to be an important risk factor for peri-implant disease (PID). The aim of this meta-analysis was to assess the association between TNF-α and IL-10 polymorphisms and PID susceptibility. The Web of Science, Cochrane Library, Scopus, and PubMed/Medline databases were searched for studies published until 12 April 2021. RevMan 5.3, CMA 2.0, SPSS 22.0, and trial sequential analysis software were used. Twelve studies were included in our analysis. The pooled ORs for the association of TNF-α (−308 G > A), IL-10 (−1082 A > G), IL-10 (−819 C > T), and IL-10 (−592 A > C) polymorphisms were 1.12, 0.93, 1.35, and 0.77 for allelic; 1.42, 0.95, 3.41, and 0.34 for homozygous; 1.19, 1.88, 1.23, and 0.49 for heterozygous, 1.53, 1.12, 1.41, and 0.39 for recessive; and 1.16, 1.87, 2.65, and 0.75 for dominant models, respectively, with all the estimates being insignificant. The results showed an association between TNF-α (−308 G > A) polymorphism and the risk of PID in patients of Asian ethnicity (OR = 1.59; p = 0.03). The present meta-analysis illustrated that TNF-α (−308 G > A), IL-10 (−1082 A > G), IL-10 (−819 C > T), and IL-10 (−592 A > C) polymorphisms were not associated with the risk of PID, whereas TNF-α (−308 G > A) polymorphism was associated with an elevated risk of PID in Asian patients.


Introduction
Despite the high survival rate and success of dental implants, it has long been known that osseointegrated implants may suffer from biological complications, collectively referred to as peri-implant disease (PID) [1]. PIDs are defined as inflammatory lesions of the tissue around the implant and include mucositis around the implant (inflammatory lesion confined to the mucosa around the implant) and peri-implantitis (an inflammatory lesion of the mucosa that affects the supporting bone with bone loss) [2]. A recent metaanalysis included peri-implantitis, implant failure, and marginal bone loss as PIDs [3]. A review study showed peri-implantitis in 28% and ≥56% of cases and in 12% and 43% of implant sites [4]. A systematic review suggested that the prevalence of peri-implantitis was approximately 22% (range: 1-47%) [5]. Another study found the prevalence of dental implant failures to be 11% in males and 9% in females; this prevalence was dependent on implant length, implant diameter, and bone quality [6]. Marginal bone loss (>0.5 mm) at implants was also recognized in 30% of cases and 16% of implant sites [7]. Evidence suggests that those who are aged more than 60 years, smokers, receiving head and neck radiation, postmenopausal, suffering from diabetes, and receiving hormone replacement therapy experienced significantly elevated implant failure in comparison with healthy patients [8].
Genetic susceptibility has been shown to be a significant risk factor for peri-implantitis, and there are numerous studies assessing this in different populations [9][10][11]. Gene polymorphisms refer to changes in DNA sequencing, such as the regulation of inflammatory mediators, primarily the gene promoter region, which can affect gene function and the progression of inflammatory diseases [12,13]. Polymorphisms of cytokines associated with the risk of PID, such as interleukin (IL)-1A [14], IL-1B [14,15], IL-6 [16,17], tumor necrosis factor-alpha (TNF-α) [17], and IL-10 [15,18] as an anti-inflammatory cytokine, could inhibit the production of proinflammatory cytokines and the induction of B lymphocyte proliferation as well as prevent the proliferation and activation of natural killer cells [19]. TNF-α is another anti-inflammatory cytokine that plays an important role in inflammatory processes [17]. The role of TNF-α in the destruction of bone around the implant has been suggested by researchers [20]. A meta-analysis [21] assessed the association of TNF-α (−308 G > A) and IL-10 (−1082 A > G) polymorphisms with the risk of implant failure by including two and three studies, respectively. Another meta-analysis [3] investigated the role of TNF-α (−308 G > A) polymorphism in PID using the data from six studies. Their results did not show any association between these polymorphisms and the risk of dental implant failure [21] and PID [3].

Design
The preferred reporting items for systematic review and meta-analysis (PRISMA) guidelines were used to report this study [22]. The PICO (patient/population, intervention, comparison, and outcomes) question was as follows: Is there an association between IL-10 and TNF-α polymorphisms and the risk of PID in patients with dental implants?

Eligibility Criteria
The studies were retrieved from the databases by one author (M.S.), and the duplicates and irrelevant studies were then excluded. The studies were considered relevant if they met the following eligibility criteria: (I) case-control design; (II) PID as the outcome of interest; (III) reporting TNF-α (−308 G > A), IL-10 (−1082 A > G), IL-10 (−819 C > T), or IL-10 (−592 A > C) polymorphisms with any genetic models; and (IV) having the required data to calculate the odds ratios (ORs) with 95% confidence intervals (CIs) for the genetic models. The studies were removed if they did not have the required data regarding genotype distributions or were animal studies, meta-analyses, review articles, letters to the editor, reported secondary data, and reported genotype distributions after treatment. The second author (L.J.) rechecked the relevant articles based on the eligibility criteria. Any disagreement between the two authors was resolved by discussion.

Data Extraction
One author (M.S.) independently extracted the data from each study and another author (J.T) rechecked them. The information retrieved from the studies included the first author's name, publication year, ethnic group, control source, mean/median age and male/female ratio in the two groups (patients and controls), genotyping method, form of disease, number of patients or controls, the p-value of Hardy-Weinberg equilibrium (HWE) in controls, the quality score, and the distribution of genotypes in the two groups. If there was a disagreement between the authors, the problem was resolved by a short discussion.

Quality of Assessment
One author (L.J) distinguished the quality of each included article using the modified Newcastle-Ottawa Quality Assessment Scale questionnaire, which was used in a similar meta-analysis involving gene polymorphisms. It involves assigning scores ranging from 0-2 and 0-1 on five (representativeness of cases, ascertainment of case outcomes, ascertainment of controls, H-W equilibrium in controls, and association assessment) and two (description of follow-up and genotyping examination) criteria, respectively. A maximum total score of 12 was possible for each study [3].

Statistical Analyses
The Review Manager 5.3 (RevMan 5.3; the Cochrane Collaboration, the Nordic Cochrane Centre, Copenhagen, Denmark) was used to calculate crude odds ratio (OR) and 95% confidence interval (CI) showing the association between IL-10 and TNF-α polymorphisms and dental PID risk in the five genetic models. To evaluate the pooled OR significance, the Z test was applied with a p < 0.05. The Cochrane Q test and I 2 statistic showed the heterogeneity (inconsistency in the polymorphism effect across primary studies). If there was a statistically significant heterogeneity (p < 0.1 or I 2 > 50%), we used a random-effect model (DerSimonian and Laird method) [23], and if there was no significant heterogeneity, a fixed-effect model (Mantel-Haenszel method) [24] was used.
The chi-square test was used to calculate the p-value of HWE in the control group of each study, with p < 0.05 indicating a deviation from the HWE.
Subgroup, sensitivity, and meta-regression analyses were performed where possible depending on the number of studies available. The subgroup analysis for explanation of heterogeneity based on a priori hypothesis was done for TNF-α (−308 G > A) polymorphism according to the ethnicity, control source, disease form, and number of individuals.
The funnel plots were analyzed by the Egger's and Begg's tests (with p-values < 0.05 indicating statistically significant existence of the publication bias). To evaluate the stability of the pooled results, we used sensitivity analyses ("one study removed" and "cumulative analysis") for TNF-α (−308 G > A) and IL-10 (−819 C > T) polymorphisms. The Comprehensive Meta-Analysis version 2.0 (CMA 2.0; Biostat Inc., Englewood, NJ, USA) was used for sensitivity analyses and assessing publication bias. Each meta-analysis may create a false-positive or -negative conclusion [25]. Hence, TSA was conducted using TSA software (version 0.9.5.10 beta) (Copenhagen Trial Unit, Centre for Clinical Intervention Research, Rigshospitalet, Copenhagen, Denmark) to reduce these statistical errors [26]. Additionally, a threshold of futility was tested by TSA to earn a conclusion of no effect before reaching the information size. The required information size (RIS) based on an alpha risk of 5%, a beta risk of 20%, and a two-sided boundary type was computed. For those analyses where the Z-curve reached the RIS line or monitored the boundary line or futility area, it was considered that the studies had adequate sample size and their results were valid. Otherwise, it was assumed that the available information was inadequate and more evidence was needed.

Study Selection
Through the electronic and manual search, 63 records were identified ( Figure 1). After removing the duplicates, 30 records were screened, while 10 irrelevant records were removed. A total of 20 full-text articles were evaluated for possible inclusion, and 8 of them were deemed irrelevant and excluded with reasons (one animal study, two reviews, one reported gingival crevicular fluid level of TNF-α and not polymorphisms, two meta-analyses, one systematic review, and one reported implant failure after total hip arthroplasty). Finally, 12 studies were included in our analysis.

Quality Assessment
The quality score for the studies based on modified the Newcastle-Ottawa Scale (NOS) is shown in Table 1. The scores ranged from 8 to 10.

Quality Assessment
The quality score for the studies based on modified the Newcastle-Ottawa Scale (NOS) is shown in Table 1. The scores ranged from 8 to 10. HB: hospital-based; PB: population-based; RT-PCR: real-time polymerase chain reaction; ARMS: amplification-refractory mutation system; RFLP: restriction fragment length polymorphism.

Subgroup Analysis
Subgroup analyses based on ethnicity, control source, disease form, and number of individuals were performed on the association between TNF-α (−308 G > A) polymorphism and PID risk ( Table 7). The results showed that ethnicity was the only significant factor. Asian patients with TNF-α (−308 G > A) polymorphism had a significant elevated risk of PID than the controls (OR = 1.59; p = 0.03), whereas there was no significant association between the polymorphism and PID risk for Caucasian and mixed ethnicities. Table 7. Subgroup analyses based on ethnicity, control source, disease form, and sample size for five genetic models of TNF-α (−308 G > A) polymorphism. N, N

Sensitivity Analysis
Sensitivity analyses were performed by removing studies with a deviation of HWE in their controls for both TNF-α (−308 G > A) and IL-10 (−819 C > T) polymorphisms (Table 8).
In addition, "one study removed" and "cumulative analyses" were performed, and the results did not change for both the polymorphisms.

Meta-Regression
To check the effect of publication year and sample size on the pooled results of TNF-α (−308 G > A) polymorphism, meta-regression was conducted. The findings demonstrated that the publication year and sample size were not confounding factors on the association between TNF-α (−308 G > A) polymorphism and susceptibility to PID (Table 9).

Trial Sequential Analysis
For TNF-α (−308 G > A) and IL-10 (−1082 A > G) polymorphisms, the Z-curve did not reach the RIS line or cross the boundary line or enter futility area, establishing that the evidence was not enough for significant results and more information was needed. With regard to IL-10 (−819 C > T) polymorphism, the Z-curve exceeded the RIS line, confirming that there was enough evidence to conclude that that the IL-10 (−819 C > T) polymorphism was not associated with the PID risk ( Figure 2).

Discussion
Dental implants provide a great treatment option for patients with missing teeth by replacing the root of the tooth with fixed permanent artificial tooth roots that are

Discussion
Dental implants provide a great treatment option for patients with missing teeth by replacing the root of the tooth with fixed permanent artificial tooth roots that are implanted into the jawbone matching the natural ones and supporting the prosthetic crowns [21].
The main results of the present meta-analysis showed that TNF-α (−308 G > A), IL-10 (−1082 A > G), IL-10 (−819 C > T), and IL-10 (−592 A > C) polymorphisms were not associated with PID risk. Out of TNF-α (−308 G > A), IL-10 (−1082 A > G), and IL-10 (−819 C > T) polymorphisms, the TSA confirmed the result of only IL-10 (−819 C > T) polymorphism, indicating the need for more evidence on TNF-α (−308 G > A) and IL-10 (−1082 A > G) polymorphisms. The TNF-α (−308 G > A) polymorphism had a significant elevated risk in Asian PID patients compared to controls. Moreover, the meta-regression confirmed that publication year and number of individuals were not confounding factors on the association between TNF-α (−308 G > A) polymorphism and PID susceptibility.
One research showed increased salivary TNF-α level in cases with peri-implant clinical condition, especially in patients with peri-implantitis [35]. Another research confirmed significantly higher serum level of TNF-α in peri-implantitis patients compared to controls, indicating the pivotal role of these cytokines in peri-implantitis [36]. Farhad et al. [37] concluded that IL-10 level increased in patients with PID compared to individuals with healthy peri-implant tissues, which was also confirmed by many other studies [38][39][40]. Differences in the level of two cytokines between PID patients and controls and the lack of association between the two polymorphisms and the risk of PID in our meta-analysis may indicate the influence of other genetic as well as environmental factors. Future studies might need to explore the influence of these factors.
A meta-analysis examined the association between smoking, radiotherapy, diabetes, and osteoporosis and the risk of dental implant failure [41]. Smoking [17,[41][42][43] and radiotherapy [41] are considered the most significant risk factors for dental implant failure. It would be interesting to explore the role of these risk factors on the relationship between gene polymorphism and PID. However, we could not run a meta-regression analysis to assess the effect of these risk factors on the association between gene polymorphisms and PID risk due to unavailability of such data. Wilson and Nunn evaluated the effect of IL-1 polymorphism (smoking and age on dental implant failures) and found that smoking was the only strong risk factor for implant failure [44]. Feloutzis et al. observed similar findings suggesting that IL-1 genotype could further precipitate the detrimental effect of smoking on peri-implant bone loss [45]. Pathogenic bacteria, lack of oral hygiene, and alcohol consumption have also been reported as factors associated with peri-implantitis [42,43]. Research has also indicated the possible effect of systemic diseases on peri-implant bone loss, and most studies therefore recruit PID patients without any systemic diseases [46][47][48][49]. Most studies in our meta-analysis selected individuals who did not smoke or the smoking status was matched between two groups (patients and controls) [14,18,27,28,30,32,33] and without any systemic disease in both cases and controls [18,27,30,33].
Although research exploring the effect of several systemic, habitual, and clinical risk factors on the risk of PID is vast, the effect of genetic risk factors has not been well studied [50,51]. This meta-analyses evaluated TNF-α (−308 G > A) and IL-10 (−1082 A > G) polymorphisms [21] or TNF-α (−308 G > A) polymorphism [3] alone, and no association was observed between any of these polymorphisms and the risk of PID disease. In our meta-analysis, there was an association between TNF-α (−308 G > A) polymorphism and PID in Asian patients. We need to further explore the role of ethnicity on the association of the mentioned polymorphisms and PID risk, especially TNF-α (−308 G > A) polymorphism.
This meta-analysis had several limitations, namely (1) few studies and lack of subgroup analysis for IL-10 polymorphisms, (2) smaller sample sizes in some of the included studies, (3) inclusion of smokers as cases and controls in some studies, and (4) the studies that included populations from Asian ethnicity were both from China, meaning the results might not be representative of all Asian population. Lack of publication bias, stability of the pooled data, and the confirmation of the pooled results by TSA would be the important strengths of this meta-analysis.

Conclusions
The pooled analysis of the present meta-analysis showed that TNF-α (−308 G > A), IL-10 (−1082 A > G), IL-10 (−819 C > T), and IL-10 (−592 A > C) polymorphisms were not associated with PID risk, whereas TNF-α (−308 G > A) polymorphism was associated with a significant elevated risk of PID in patients of Asian ethnicity. Funding: No financial support was received for this study.

Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.

Data Availability Statement:
The data that support the findings of this study are available on request from the corresponding author.