A Cross-Sectional Analysis on Factors Associated With Peri-Implant Pathologies at the Implant Level Open Access

Dental implants have become a good strategy for replacing missing teeth in patients. However, biological complications such as peri-implantitis and peri-implant mucositis are more frequent during the late stages of implant rehabilitation, potentially leading to implant failure.^1,2  Several reports indicate that these pathologies are a common finding in such patients.^3–6  The occurrence of peri-implantitis varies between 8.9% and 56% in subjects and/or dental implants while mucositis presents a frequency of 43% to 80%.^3–7  Divergences regarding this prevalence may be due to a number of factors, including the diagnostic criteria employed and the characteristics of the population evaluated.

In addition to the dental biofilm, some factors have been associated with the onset and progression of peri-implant pathologies.^3,6,8,9  Periodontal disease, smoking, lack of supportive periodontal therapy, diabetes, the type of rehabilitation, and parafunctional habits can all affect the incidence of peri-implant disease.^1,7,10,11  Nevertheless, additional epidemiological and clinical studies are required to investigate the biological and local factors that contribute to peri-implant pathologies, to provide robust scientific evidence regarding disease etiology.¹² 

Studies of the possible impacts of systemic and local factors on the occurrence of peri-implant pathologies are important to optimize the long-term prognosis of dental implant treatment. Therefore, this study aimed to investigate the occurrence of peri-implant pathologies in dental implants, as well as possible factors associated.

This cross-sectional study (STROBE checklist was used; STROBE stands for an international, collaborative initiative of epidemiologists, methodologists, statisticians, researchers, and journal editors involved in the conduct and dissemination of observational studies, with the common aim of STrengthening the Reporting of OBservational studies in Epidemiology; Supplement 1) was previously approved by the Research Ethics Committee of the Federal University of Rio Grande do Norte, Brazil (protocol: 186/10-P). The patients were selected in the period from 2000 to 2011.

Patients undergoing periodontal therapy or on antibiotic therapy during the 2 months prior to the clinical examination were excluded. Individuals that had received implant prostheses during the last 6 months were included.

The medical and dental histories of all patients were updated upon inclusion in the study. Parameters were recorded as follows: Probing depth (PD) measured in mm at 6 sites per implant and teeth; full-mouth visible plaque index (VPI);¹³  full gingival bleeding index (GBI);¹³  bleeding on probing (BOP), described as presence (1) or absence (0) at 6 sites per teeth and implants; presence of suppuration (S), described as presence (1) or absence (0); keratinized mucosa (KM), described as presence (1) or absence (0). The bleeding and plaque score indexes were divided into 2 categories as follows: 0–30% and >30%. A professional performed all clinical examinations using a WHO (World Health Organization) periodontal probe (PD: ICC [intraclass correlation coefficient] = 0.89 and KM: ICC = 1.00) and a conventional millimeter periodontal probe (PCP-UNC 15; Hu-Friedy).

Periapical radiograph was used for the evaluation of the implants (paralleling and positioners). The height of the bone was evaluated (length of exposed implant threads). Characteristics established by the American Academy of Periodontology¹⁴  were used to classify periodontal disease. Bleeding on probing was present with no attachment loss for gingivitis. Probing pocket depth was >4 mm, the clinical attachment level was ≥3 mm and bleeding on probing classified periodontitis.

The references for peri-implant diagnosis were based on those of a study by Mombelli et al¹⁵ Bleeding on probing and/or gingival and/or suppuration, with no radiographic bone loss were employed to diagnose mucositis, while PD ≥ 5 mm and/or suppuration, and radiographically visible bone loss were used to identify peri-implantitis. Estimating a real proportion of 0.4 and a ratio of the null hypothesis of 0.6 and significance level of 0.05, this sample presented a power of 99.99%.

Data were statistically analyzed using the SPSS software (SPSS Inc), and the implant was used as a statistical unit. All statistics, as well as all results and conclusions, were reviewed by a person with experience in the field.

To verify significant associations, the chi-square test was used. Associations with a value of P < .20 were selected for multiple logistic regression analysis (hierarchical method), which allows us to estimate a probability associated with the occurrence of a given event in a set of explanatory variables. This method was used to insert variables in the model even when no statistically significant differences were detected (dependent on the model adaptation capacity, according to the Hosmer and Lemeshow model). To measure the effect of multiple analysis, the adjusted odds ratio was transformed into the adjusted prevalence ratio, with the objective of avoiding overestimation of the effect. The level of significance established for all statistical tests was 5%.

A total 523 implants and 2,718 teeth were evaluated; 79.4% (n = 123) of the subjects were women and the average age of patients was 54.05 years (±12.61; minimum: 18 years; maximum: 88 years). Systemic changes were reported for 34.2% (n = 53) of patients and dry mouth feeling for 17.4% (n = 27). Smokers totaled 12.9% (n = 20) of patients and 36.1% (n = 56) drank alcoholic drinks. Table 1 presents the descriptive analysis of data.

The teeth count in the mouth ranged from zero to 31 and the highest amount of implants per patient was 11. The mean full-mouth VPI was 48.83% (±26.60) and the mean GBI was 8% (±13.71). The average time of prosthesis use was 47.42 months (±26.45; minimum: 6; maximum: 120 months) (see Table 1).

Periodontitis was observed in 50% of patients, gingivitis was detected in 43% and 7% presented healthy tissues. Five hundred and twenty-five implants were evaluated, of these 56.6% (n = 296) presented peri-implant pathology and 43.4% were healthy (n = 227); diagnoses, according to the amount of implants, are presented in Table 1. Associations between the dependent variable, “peri-implant pathology,” and demographic independent variables also related to dental implants are presented in Table 2. Associations between the dependent variable and prosthetic variables are presented in Table 3, and associations between the dependent variable and clinical variables are shown in Table 4.

The univariate analysis showed that variables, such as use of medications (yes), systemic alterations (yes), time of prosthesis use (>2 years), location of the implant (upper jaw), region (posterior), accomplishment of previous reconstructive surgery (such as augmentation procedure) (yes), GBI (>30%), and keratinized mucosa (absence) were associated more with the occurrence of peri-implant pathologies (Tables 2 and 4).

The binary logistic regression indicated that the predictive variables of peri-implant pathology; GBI of >30%, >2 years of prostheses use, upper jaw location, and use of medications were independent factors that were associated with peri-implant pathology, when adjusted by augmentation procedure and sex (Table 5).

Recent reports have shown frequencies of peri-implantitis and mucositis of 43.2%⁷  and 45.6% in patients with implants.¹⁶  A meta-analysis by Lee et al (2017)¹⁷  found a mean prevalence of implant-based peri-implantitis of 9.25% (95% confidence interval [CI]: 7.57–0.93) and a mean implant-based mucositis prevalence of 29.48% (CI: 22.65–36.32). These pathologies are present in elevated frequencies and associated with the lack of a regular maintenance.

The variability in the reports of peri-implantitis may be explained by the fact that different clinical parameters are used to evaluate and define this pathology; furthermore, the lack of standardization of these parameters makes comparisons among reports difficult.¹⁸  Despite recent publications classifying periodontal and peri-implant pathologies and conditions,^7,19,20  the standardization of these clinical parameters, especially when peri-implantitis is evaluated, is still controversial, as variations in the depth of probing may not be confirmed by radiographic examination. It should also be noted that some implant platforms undergo a physiological process of bone loss that does not necessarily constitute disease.

Plaque control is indispensable, as peri-implant pathologies are of infectious origin.^21,22  However, the contribution of plaque accumulation to the increased risk of peri-implant pathologies is not clear. A VPI > 30% was categorized in this study with the purpose of considering generalized and localized poor plaque control. Despite the failure to demonstrate a clear association between the presence of full-mouth plaque (P = .298; VPI) and peri-implant pathologies, in increased risk of susceptibility, other studies have made similar assessments, where VPI > 25% presented an association with peri-implant pathology.^2,7 

Furthermore, statistically significant differences in terms of a higher percentage of GBI (P = .005) may justify this parameter as a predictive variable for the occurrence of peri-implant pathology, as a GBI of greater than 30% was observed (PR = 1.496). Some studies have pointed out a positive correlation between high plaque scores and an increased risk of presenting peri-implant pathology.^17,23  Therefore, the primary role of plaque was highlighted as a dose-dependent association between higher plaque scores and worse conditions in peri-implant pathologies.³  Gingival bleeding may reflect the represent of biofilm accumulation on the implants reinforces the role of this factor of risk.¹¹ 

Whereas changes in volume, such as edema, and color of marginal peri-implant tissues were considered as signs of diseases.^18,24  The absence of keratinized tissue, or the presence of scar tissue from previous surgeries, may influence the dental researcher at the time of quantifying the degree of inflammation by visual criteria. The inability to reproduce these indices has been cited in some other studies.^18,25,26 

The keratinized mucosa has been identified as a protective factor against the incidence of peri-implant pathologies, and grafting procedures for keratinized mucosa gain are an important factor for reducing indices such as bleeding and maintenance of bone tissue.²⁷  More recent studies have reported that the absence of keratinized mucosa may hamper oral hygiene procedures and have shown limited evidence of the association of this characteristic with peri-implant pathologies.^7,20  The univariate analysis used in this study showed that the absence of keratinized mucosa was associated more with the incidence of peri-implant pathologies; however, in the multivariate analysis, this variable strength was lost within the context of this study.

Other possible factors associated with peri-implant pathologies have been reported; this study observed more events in men, but this difference was not statistically significant, as also observed by other studies.^{3,6,7,10,28,29}  Women display better oral hygiene, which can improve health³⁰  and may explain the higher prevalence of disease in males in this study. It should be noted that patients in this study presented with a history of periodontal disease and, therefore, this limitation may not have the power to correlate the associated factors investigated or correctly evaluate the effect and cause relationship.

Systemic changes that could influence peri-implant health (osteoporosis, immunocompromised patients, diabetes, cardiovascular disorders, and hormonal disorders) were also evaluated in this study and were found to be associated more with peri-implant pathologies (P = .016), as well as the use of medications. As these two variables compete between themselves, the use of medications was chosen for application in the model of regression, since it demonstrated the best fit to the model. Indeed, the use of medication was an independent factor associated with the occurrence of peri-implant pathologies (PR: 1,261).

While, Roos-Jansåker et al¹  did not find any significant associations among some of these conditions (osteoporosis, diabetes, and coronary heart disease) and the onset of disease. The medications in use that may have an effect on bone turnover and/or healing and remodeling of mucosa could have affected the results.¹  This aspect was compromised, since the medication type was not considered separately in the anamnesis of this study for additional discussion.^6,16 

Periodontal and peri-implant pathologies were not significantly associated with each other, as also observed in other studies.^1,7,11,28  Microorganisms remaining in periodontal pockets can be transferred from to peri-implant sites, ie, pockets as niches of infection.^8,31  An elevated incidence of biological complications associated with implants has been observed in patients with periodontitis, compared to healthy individuals.^32,33  Additionally, periodontitis as a risk indicator may be attributable to the fact that both pathologies share common host factors or common microbiota.⁹ 

The finding that peri-implant pathologies are positively correlated with the time of prostheses use⁵  has also been confirmed by the present analysis, as peri-implant pathology was more associated with implants that had been fitted more than 2 years before (PR: 1,720). Similar findings were observed in other studies.^7,17,34  It is important also to note that the time of use of the prosthesis, when the prosthesis is well planned and executed, should not be an important factor in the predictability of peri-implant pathology incidence. Other factors such as occlusion, prosthetic condition, adaptation, type, material, fixation (excess cement), and complexity of rehabilitation have been more frequently reported;^7,17,35,36  however, in our study, these factors were not important predictors, possibly due to the fact that our unit of evaluation was the implant itself.

The rehabilitation of dental implants in grafted areas, even though this is still considered a challenging practice, is currently regarded to be a successful procedure, independent of the material used (autogenous, allograft, xenograft, and alloplast). Failures in the grafted regions are more associated with contamination of the material, lack of primary stability, lack of installation technique and inadequate prosthetic rehabilitation, as well as disrespect to biomechanical principles.³⁷  It should be noted that the type of increase (horizontal or vertical), extension of the grafted area, the number of implants, and countless other variables can influence the success of the rehabilitation.³⁷  However, the authors conclude that the type of platform, smoking, and history of periodontitis are more important in this context.³⁸  Reports of peri-implant pathologies occurring in association with the grafted area are scarce in the literature and additional studies are needed to elucidate this issue. We, herein, observed a significant association between peri-implant pathologies and grafting area, but this variable lost strength in the model; nevertheless, this association was still apparent when adjusting the model and, consequently, the magnitude of the strength of the other variables. Several authors have observed a higher occurrence of peri-implant pathology in the upper jaw.^1,7,10,11,12  This study also showed an association between location in the upper jaw and peri-implant pathology. Serino and Turri³⁹  demonstrated that bone characteristics and quantity may influence the progression of peri-implantitis, especially in areas with smaller bone density and lower density, which are probably more susceptible to the inflammatory process and occlusal stresses.

Even in the case of a cross-sectional study, with the limitations inherent to this type of study (evaluation of the specific moment), the factors associated with peri-implant pathologies are important to highlight. Potential risk factors should be confirmed with specific studies for this type of analysis; as such the present study represents an important starting point for the development of new research in the area. The frequency of peri-implant pathologies in implants was 57.2%. The most common factors were: more than 2 years of prosthesis use, location in upper jaw, GBI > 30%, and the use of medications, when adjusted by augmentation procedure and sex.

BOP:

bleeding on probing

CI:

confidence interval

GBI:

full gingival bleeding index

KM:

keratinized mucosa

NA:

not applicable

NI:

not informed

PD:

probing depth

PR:

prevalence ratio

S:

presence of suppuration

VPI:

full-mouth visible plaque index

WHO:

World Health Organization

The authors declare no conflict of interest.