Clinical research on peri-implant diseases: consensus report of Working Group 4

Authors

  • Mariano Sanz,

    Corresponding author
    • University Complutense of Madrid – Periodontology, Madrid, Spain
    Search for more papers by this author
    • Conflict of interest and source of funding statement

  • Iain L. Chapple,

    1. Periodontal Research Group, The University of Birmingham, Birmingham, United Kingdom
    Search for more papers by this author
    • Conflict of interest and source of funding statement

  • on behalf of Working Group 4 of the VIII European Workshop on Periodontology*


  • The authors declare that they have no conflict of interests. This workshop was financially supported by the European Federation of Periodontology and by unrestricted grants from Astra, Nobel Biocare and Straumann.

Address:

Mariano Sanz

Universidad Complutense de Madrid

Facultad de Odontologia

Plaza Ramon y Cajal

E-28040 Madrid

Spain

E-mail: marianosanz@odon.ucm.es

Abstract

Background

Two systematic reviews have evaluated the quality of research and reporting of observational studies investigating the prevalence of, the incidence of and the risk factors for peri-implant diseases and of experimental clinical studies evaluating the efficacy of preventive and therapeutic interventions.

Materials and Methods

For the improvement of the quality of reporting for both observational and experimental studies, the STROBE and the Modified CONSORT recommendations were encouraged.

Results

To improve the quality of research in peri-implant diseases, the following were recommended: the use of unequivocal case definitions; the expression of outcomes at the subject rather than the implant level; the implementation of study validation tools; the reporting of potential sources of bias; and the use of appropriate statistical methods.

Conclusions

In observational studies, case definitions for peri-implantitis were agreed. For risk factor determination, the progressive use of cross-sectional and case–control studies (univariate analyses), to prospective cohorts (multilevel modelling for confounding), and ultimately to intervention studies were recommended. For preventive and interventional studies of peri-implant disease management, parallel arm RCTs of at least 6-months were encouraged. For studies of non-surgical and surgical management of peri-implantitis, the use of a composite therapeutic end point was advocated. The development of standard control therapies was deemed essential.

Introduction

  • *Working Group 4:
  • Mariano Sanz,
  • Iain Chapple,
  • Jan Derks,
  • E. Figuero,
  • Filippo Grazianni,
  • Lisa Heitz-Mayfield,
  • David Herrera,
  • Ann-Marie Jansaker,
  • SÖ ren Jepsen,
  • Bjorn Klinge,
  • Bruno Loos,
  • Andrea Mombelli,
  • Panos Papapanou,
  • Ioannis Polyzois,
  • Stefan Renvert,
  • Giovanni Salvi,
  • Industry representation:
  • Pascal Kunz,
  • Anna-Karin Lundgren,
  • Rene Willi.
image

The remit of this working group was to provide answers to key aspects of clinical research into peri-implant diseases focusing on aspects of study design, reporting and outcome measurements.

Two reviews had evaluated the quality of research and the quality of reporting of observational studies investigating the prevalence of, the incidence of and the risk factors for peri-implant diseases (Tomasi & Derks 2012), as well as those experimental clinical studies evaluating the efficacy of preventive and therapeutic interventions (Graziani et al. 2012) in a systematic manner.

Quality of Reporting, Case Definitions and Methods to Study Incidence of, Prevalence of and Risk Factors for Peri-Implant Diseases

The systematic review evaluating the quality of reporting of the observational studies investigating the incidence of, prevalence of and risk factors for peri-implant diseases (Tomasi & Derks 2012) highlighted that many studies have solely reported implant-based data analysis or have not provided a clear definition of the peri-implant diseases being studied.

Another important issue emphasized in the review was the different criteria used for defining a “case” in studies investigating the prevalence of peri-implant diseases. Whilst studies performed since the previous European Workshops have used the definitions developed in those consensus meetings (Lang & Berglundh 2011), there are still differences in the thresholds employed for radiographic bone loss used to define peri-implantitis and the reference time point from which this bone loss has occurred.

The selected studies investigating the prevalence of peri-implant diseases have used cross-sectional study designs and convenience samples, mostly of limited size, with clinical and radiological outcomes as the basis for defining health, versus peri-implant mucositis and peri-implantitis. These convenience samples may not be representative of the target population.

When analysing the quality of reporting in the selected studies, the STROBE recommendations (Vandenbroucke et al. 2007a, b, c) have been utilized. Although most of the studies lack reporting of the possible sources of bias and few have evaluated aspects of internal validity, approximately 60–70% of the recommendations were fulfilled.

From the outset, the group agreed and acknowledged the strategy of the authors to focus on investigations assessing prevalence of those peri-implant diseases based at the subject level. The group considered that the impact of peri-implant diseases upon individuals was the outcome of interest, rather than the impact upon individual implants.

When discussing the different case definitions employed in the studies evaluating prevalence and incidence of peri-implant diseases, the group agreed that the definitions established in the previous EWP consensus workshops published in 2008 (Lindhe & Meyle 2008) and 2011 (Lang & Berglundh 2011) should be adopted. This consensus attempts to further clarify some key aspects that are relevant to the conduct of observational studies to improve the methodological quality and reporting standards of clinical research in peri-implant diseases.

For defining a case as peri-implantitis, it was agreed that the components described in previous consensus reports must be present: “changes in the level of crestal bone, presence of bleeding on probing and/or suppuration; with or without concomitant deepening of peri-implant pockets” (Lang & Berglundh 2011). It was strongly recommended that for investigations assessing the prevalence and incidence of peri-implant diseases, all studies must be underpinned by basic quality assurance procedures including internal validation of examiner reproducibility and the use of measurement instruments aimed at reducing their inherent variability. It is recommended that long-cone parallel radiographic projection techniques were utilized to assess the crestal bone changes. Details of the probing protocol utilized should also be described.

In this context, the group agreed that different approaches should be used when dealing with studies of disease prevalence and incidence.

Studies on disease prevalence

For disease prevalence, it was accepted that individuals presenting with peri-implant diseases might not have baseline clinical or radiological measures available to benchmark current findings, and therefore more pragmatic case definitions should be developed to serve as the basis for diagnosis of peri-implantitis and on clinical decision making for therapeutic purposes . Here, the focus is specificity of diagnosis rather than sensitivity, accepting the possibility of false negative diagnoses. In the absence of previous radiographic records, a threshold vertical distance of 2 mm from the expected marginal bone level following remodelling post-implant placement is recommended, provided peri-implant inflammation is evident. If previous radiographs are available, a more sensitive threshold for measurable bone loss can be used depending on the reproducibility of the radiological assessment method employed.

Studies on peri-implantitis incidence

For prospective studies evaluating the incidence and or progression/recurrence of peri-implantitis, baseline clinical measures and radiographs are essential. Since these studies must be planned thoroughly and ethical approval must be obtained, baseline clinical and radiological data should be established once the remodelling phase post-implant placement has occurred. In this manner, in the presence of inflammation, any detectable bone loss beyond the inherent variability of the radiological measurement error, should be considered as loss of supporting bone and hence, used in the case definition. Such reference data permits more sensitive case definitions to be developed, accounting for measurement errors, but providing higher sensitivity and accepting the potential for some false positive assignments of peri-implant disease. For these incidence studies, a threshold of detectable bone loss of 2–3 times the SD of the measurement error is recommended (1.0–1.5 mm).

Studies on risk factors for peri-implant diseases

Studies addressing risk factors in peri-implant diseases are in their infancy. Traditionally, risk factors are identified on the basis of a multi-step process including:

  • Identification of putative risk factors using cross-sectional or case-control studies and univariate analyses, followed by multi-variate analyses aimed at identifying potential confounding factors
  • Studies validating these putative factors, which employ a prospective cohort design, ideally confirmed in different populations
  • In the case of modifiable true risk factors, intervention studies demonstrating lower incidence of disease following targeted control of the respective factors

The currently available literature primarily consists of cross-sectional and case–control studies, which are limited in their generalizability. Based upon the approach taken in this systematic review (Tomasi & Derks 2012), four prospective cohort studies exist with limited sample sizes, and based upon convenience, samples rather than broader sampling of the population exist . There is heterogeneity in the risk indicators investigated across the broad categories of host-derived, lifestyle, environmental and local factors. Given the limited sample sizes investigated so far, it is not possible to assess with precision the interactions between the different exposures of interest.

Recommendations for future research

Based upon the available evidence, the group recommends that additional and adequately powered research should be conducted to validate the most plausible putative risk factors for the onset and progression of peri-implant diseases, including:

  • History of treated and/or current periodontitis
  • Oral hygiene
  • Smoking and other subject-related factors
  • Local factors such as malpositioning of implants (within the basal bone, inter-implant, implant-tooth), lack of cleansability of the reconstruction, excess cement and implant surface characteristics

To establish valid and reliable statistics on the prevalence and incidence of peri-implant diseases, it is recommended that national/regional databases of patients receiving implant therapy be established. These should include both baseline and annual follow-up assessments of implant status and potential exposures.

These registries will facilitate the conduct of future studies of risk factor assessment that are:

  • Adequately powered and adjusted
  • Representative of the population at large
  • Of sufficient duration
  • Inclusive of different settings (University, Specialist Practice, General Practice)

In the absence of such databases, it is recommended that multicentre studies be conducted to enhance the generalizability (size, power) of study outcomes.

Study quality will be improved by:

  • Establishing unequivocal case definitions
  • Involving calibrated examiners
  • Careful assessment and reporting of exposures (e.g. oral hygiene status, periodontal status and smoking history)
  • Selection of unbiased samples of sufficient size
  • Provisions to account for loss to follow-up
  • Appropriate statistical analyses

Employing validated tools such as the STROBE guidelines will attain improvements in the quality of reporting.

Quality of Reporting Outcome Measurements and Methods to Study the Efficacy of Preventive and Therapeutic Approaches to Peri-Implant Diseases

The systematic review evaluating the efficacy of preventive and therapeutic approaches to peri-implant diseases (Graziani et al. 2012) focussed upon the selection of randomized clinical trials (RCTs) and controlled clinical trials (CCTs) assessing the efficacy of preventive measures to preserve peri-implant health and therapeutic interventions aimed at treating per-implant mucositis and peri-implantitis either non-surgically or surgically.

Preventive studies of peri-implant diseases

There was a notable paucity of clinical trials addressing the efficacy of preventive interventions. From 155 studies identified as potentially fulfilling the review criteria, only seven were selected. Studies were not considered either because their experimental design was not an RCT or a CCT, or because the study evaluation time was too short. The selected studies utilized different experimental designs (parallel, cross-over and split mouth) and usually employed small sample sizes and short evaluation times. The interventions assessed included the use of adjunctive antimicrobials or specific preventive interventions aimed at plaque control around the implant supported restorations, compared with the patient's self-performed plaque control and with professional supportive therapy. Reported outcomes employed to evaluate the efficacy of the tested interventions have mainly been the use of bleeding and plaque indices, although in some studies, additional clinical (changes in PPD) and microbiological outcomes were used. The studies demonstrated moderate adherence to modified CONSORT recommendations (<50% of the items) and lacked reporting of potential sources of bias.

A significant problem was identified with regard to the lack of a standard preventive measure with demonstrated efficacy to preserve peri-implant health, which could be used as a standard control treatment in future intervention trials. It is recommended that with respect to experimental design, the 6-month parallel arm RCT should be employed; comparing the tested preventive interventions with a combination of the patient's routine self-performed plaque control measures and regular professionally delivered supportive therapy (oral prophylaxis). The objective of therapy should be the absence of mucosal inflammation around the functioning dental implants, and hence the primary outcomes should be the evaluation of the changes in mucosal inflammation (e.g. the modified bleeding index) and the absence of bleeding upon probing.

Treatment of peri-implant mucositis

There were only six parallel arm RCTs evaluating the adjunctive effect of antimicrobial compounds (Chlorhexidine, Triclosan and Essential Oils) in the treatment of peri-implant mucositis. These studies have used short evaluation times (3–8 months) and share the problems of small sample sizes, the lack of a clear definition of the problem investigated (peri-implant mucositis) and incomplete registration of periodontal status in the studied samples. As outcome measures, the reported studies evaluated the inflammatory status of the peri-implant mucosa (BOP), probing depths and plaque indices. Three studies assessed microbiological outcomes and one reported the percentage of lesions resolved. When analysing the quality of reporting and performance in the selected studies, there was between 17% and 90% adherence to the modified CONSORT recommendations.

The group agreed that the parallel arm RCT of at least 6-months duration (and including 3-month examinations) should be used for evaluating therapies to treat peri-implant mucositis. The endpoint of therapy should be the resolution of peri-implant mucosal inflammation (frequency distribution of resolved lesions) as determined by the absence of bleeding upon probing. As secondary outcomes, probing pocket depth reductions and outcomes assessing host–parasite interactions (presence of inflammatory biomarkers in peri-implant fluid and/or microbiological assessment of subgingival plaque samples) may be employed. Information on the subject's characteristics (e.g. smoking), the impact of the periodontal status and the status of the reconstruction (access for plaque control, etc.) should also be reported.

Non-surgical treatment of peri-implantitis

Six parallel arm RCTs and one split-mouth study have evaluated the efficacy of non-surgical therapies for the treatment of peri-implantitis. These intervention studies have used different case definitions for peri-implantitis and have small sample sizes, along with short evaluation periods (4.5 months; only one 12-month study) and a lack of a clear description of the periodontal status of the sample studied. There was no standard control treatment in these studies; however, mechanical debridement was included in all treatment arms. The tested interventions were either the adjunctive use of local antibiotics or alternative implant surface debridement methods (laser, ultrasonic or air abrasive devices). The outcomes utilized in these studies were reductions in PPD, reductions in the sites bleeding (BOP) and reductions in clinical attachment levels (CAL), and three studies also evaluated microbiological and radiological outcomes and one reported suppuration. A common finding in the reported studies was reductions in PPDs and BOP following the test and control interventions; however, it remains unclear as to how frequently the end point of therapy (resolution of inflammation and shallow probing depths) was met, as it was only reported in one study.

The group agreed that the lack of a standard control mode of non-surgical therapy to treat peri-implantitis was problematic. At the present time, there are no data indicating that peri-implantitis lesion severity can be the basis for recommending non-surgical or surgical therapy. As an RCT evaluating non-surgical therapy versus an untreated control arm is unethical, case series and prospective cohort studies may be used to better define the effect of non-surgical therapy alone. The parallel arm RCT with a mechanical treatment control arm should be utilized to evaluate adjunctive/alternative therapies. These studies should include both short-term (1–3 months) and longer-term evaluation times (6 and 12-months). At 3-, 6- and 12-months, probing pocket depth, bleeding on probing and suppuration should be assessed. In addition, the maintenance of bone levels should be assessed radiologically at 12-months. It is recommended that a composite outcome of disease resolution is included (absence of deep probing pocket depths with bleeding and suppuration). As secondary outcome measures, inflammatory biomarkers in peri-implant fluid and/or microbiological assessment of sub-mucosal plaque samples may be used. Information on the periodontal status of the remaining dentition and on patient reported outcomes (smoking, discomfort, aesthetic consequences, etc.) should be reported. Multi-centre approaches are advocated to achieve sufficiently powered studies.

Surgical treatment of peri-implantitis

There are three parallel arm RCTs and three CCTs that have evaluated the efficacy of surgical therapies for the treatment of peri-implantitis. These intervention studies have used different case definitions for peri-implantitis, and have in common, small sample sizes and the lack of a clear description of the periodontal and smoking status of studied sample. There was no standard control intervention in these studies; however, access flap, including debridement/degranulation of the lesion and decontamination of the implant surface was included in all treatment arms, with some studies adding systemic antibiotics and others adding only adjunctive antimicrobials (Chlorhexidine). As interventions, some had employed different modes of decontaminating the implant surface with laser devices, air abrasives and re-shaping the titanium surface, whereas other therapies have attempted to reduce the intra-bony defect, either by resection or by treating the defect with biomaterials and/or membranes. The paucity of published clinical trials, the limited sample sizes (n = 17–38) and the heterogeneity of treatments tested prevent the drawing of definitive conclusions on the efficacy of these interventions. The evaluation of reporting according to the modified CONSORT recommendations resulted in <60% adherence in four of the studies.

It is recommended that a standard mode of surgical therapy is identified for the treatment of peri-implantitis is identified. This therapy should include a clear surgical design, a proven method of decontaminating the implant surface and an appropriate means of infection control. From the six selected clinical trials, only two have used adjunctive systemic antimicrobials as part of the surgical therapy regime, whereas the other four have only used adjunctive antiseptics. There is therefore a lack of clear scientific evidence whether the adjunctive use of systemic antimicrobials might be recommended as part of the standard mode of therapy. An RCT testing this hypothesis is needed.

There is a consensus that for the evaluation of different surgical therapies, the parallel arm RCT should be used, with assessment of the end points of the therapy at least at 6 and 12 months.

It is recommended that a composite outcome of disease resolution is included (absence of deep probing pocket depths with bleeding and suppuration and no additional bone loss).

As principal outcome measurements, resolution of mucosal inflammation, reduction in probing pocket depths and changes in the bone levels should be employed. As secondary outcomes, levels of inflammatory biomarkers in peri-implant fluid and/or microbiological assessment of sub-marginal plaque samples may be used. Information on the periodontal status of the studied samples and on patient reported outcomes (discomfort, aesthetic consequences, etc.) should be reported.

It is strongly recommended that authors adhere to the reporting guidelines, as detailed in the systematic review (Graziani et al. 2012). Multi-centre approaches are encouraged to achieve sufficiently powered studies and representative populations of diverse defect morphologies and severities.

Ancillary