Microbiological findings in early and late implant loss: an observational clinical cohort study

doi:10.21203/rs.3.rs-58719/v2

Background

Implants are a predictable and well-established treatment method in dentistry. Nevertheless, early and late loss must be distinguished when examining possible failures of dental implants. The aim of the study was to report microbiological findings on the surface of implants with severe peri-implantitis, which had to be explanted.

Methods

Fifty-three specimens of implants from 48 patients without severe general illnesses were examined. The groups investigated were implants that had to be removed in the period of osseointegration (early loss, 13 patients with 14 implants) or after the healing period (late loss, 14 patients with 17 implants). Implant losses were compared with two control groups (implants with no bone loss directly after complete osseointegration, two to four months after implant placement (17 patients with 17 implants) and implants with no bone loss and prosthetic restoration for more than three years (5 patients with 5 implants)). Data for the bacteria located in the peri-implant sulcus were collected using amplification and high-throughput sequencing of the 16S rRNA gene.

Results

The biofilm composition differed substantially between individuals. In both early and late implant loss, Fusobacterium nucleatum and Porphyromonas gingivalis were found to be abundant. Late-loss implants showed higher bacterial diversity and higher abundances of Treponema, Fretibacterium, Pseudoramibacter and Desulfobulbus, while microbial communities of early-loss implants were very heterogeneous and showed no significantly more abundant bacterial taxa.

Conclusions

Specific peri-implant pathogens were found around implants that were lost after a primarily uneventful osseointegration. P. gingivalis and F. nucleatum frequently colonized the implant in early and late losses and may therefore be a general characteristic of implant loss. The reasons for early losses seem to be multifactorial, with bacterial infection being one potential risk factor, while peri-implant infections are known to be a major risk among other aspects, such as overload or mechanical complications.

Dentistry

Head & Neck Surgery

dental implant

peri-implantitis

early implant loss

late implant loss loss

microbial biofilm

16s rRNA

bacteria

Dental implants have become a routine and predictable surgical procedure in recent decades, with high rates of osseointegration and long-term success (1). Since the success of early implantology, this field has become established in daily treatment.

The reasons for tooth loss are multi-causal. Frequent problems are the development of caries at the margins of crowns, pulpal infections, periodontitis and mechanical complications, such as fractures of teeth (2-5). Although the first two factors do not affect dental implants and the rates of mechanical complications of the implant itself are low, peri-implant infections are a major risk factor for implant failure (6). Early and late explantation can be distinguished by the point in time when implants are lost.

Early implant loss occurs before prosthetic loading (7). In those cases, osseointegration is not successful; the implant is surrounded by connective tissue and cannot be used to anchor the planned denture. Some authors state that the reasons for this event may be bacterial infection, surgical mistakes such as inadequate cooling during implant bed preparation or overload of the implant by a provisional denture (7). The only treatment option is to remove the implant and repeat the surgical procedure, which is quite frustrating for both the patient and the surgeon. On the other hand, primary osseointegrated implants can lose their connection to the surrounding bone (6), and bacterial infections of the peri-implant tissues have been reported in several investigations, similar to the pathogenesis of periodontitis, as a causative factor (8). The infection leads to an inflammatory response, causing peri-implant mucositis. If this soft tissue reaction is not controlled, the infection will lead to an inflammatory process of the peri-implant bone, resulting in bone resorption. This plaque-associated bone loss is defined as peri-implantitis (9). Several health conditions, such as periodontitis, diabetes or smoking, have been discussed as risk factors for peri-implantitis (8, 10, 11).

In recent decades, knowledge of bacterial infections, especially biofilms, has fundamentally changed (12, 13). Although it has been suggested in several previous studies based on genetic analysis methods that pathogens in periodontitis and peri-implantitis are similar (14-16), it is well established by new and more accurate diagnostic methods that there are clear differences regarding the microbiota in those diseases (17-22). Additionally, the individual reaction of the immune system as well as changes in the bacterial constellation around titanium implants have been discussed (23, 24). On the other hand, some authors even state that qualitative findings in pockets of infected implants are similar to those in healthy sites, questioning if the number of bacteria is the main reason leading to the disease (25). New findings may either support or question the knowledge of bacterial findings around dental implants. Furthermore, there are no data regarding the bacterial profiles of early-loss implants, emphasizing the need for this study.

The aim of this study was to analyse the bacterial population in the peri-implant pocket of implants with severe peri-implantitis and to compare the bacterial biofilms of implants with early and late loss.

The present clinical study investigated 53 dental implants from 48 patients. The main focus of the study was a microbial comparison of findings on the surface of implants that were lost during the time of osseointegration (early loss) and implant loss because of peri-implantitis (late loss). All results were collated by the same dentist at the Dental Academy for Continuing Professional Development (Karlsruhe, Germany).

General inclusion criteria were implants with severe loss of bone-to-implant contact, either in the period of healing after insertion (< 3 months, “early loss”) or after uneventful osseointegration and loading (> 3 years, “late loss”). The exclusion criteria were as follows: patients younger than 18 years; radiation or bisphosphonate therapy; patients with untreated periodontitis and severe general conditions, such as uncontrolled diabetes (HbA_1c > 58 mmol/mol/> 7,5%); tumours; severe heart disease or a reduced state of general health. Patients who had taken antibiotics within the last two months prior to sample collection were excluded. Smokers were not excluded from the study. Only patients with one or more implants that had been inserted in our clinic were included. Only implants with severe peri-implant bone loss and no chance of regeneration or preservation were included. Implants that could not be preserved due to severe peri-implantitis were defined as having increased peri-implant pocket depth in combination with bleeding on probing, persistent symptoms and radiological bone loss of at least 6 mm. Additionally, specimens from implants without peri-implantitis (no bone loss) were integrated into the study as a control group.

Implants with peri-implantitis and bone loss but chances of preservation were excluded.

Study population

The patients were divided into four groups:

Twenty-seven patients with severe peri-implantitis and implants without a chance of preservation were divided into two groups.

Group E: early implant loss (implants with severe peri-implantitis during osseointegration prior to prosthetic restoration, ≤ 3 months after implant placement), 13 patients with 14 implants
Group L: late implant loss (implants with severe peri-implantitis and prosthetic restoration for more than three years), 14 patients with 17 implants (three patients with two implants)

We created two control groups for comparison to the healthy oral situation:

Group CE: control group (implants with no bone loss directly after completed osseointegration, two to four months after implant placement), 17 patients with 17 implants
Group CL: control group (implants with no bone loss and prosthetic restoration for more than three years), 5 patients with 5 implants

Note that one patient contributed two implants to the late-loss group (L) and an additional implant to the CE group, effectively being counted twice in the number of patients per group.

The implants were placed by three different surgeons with several years of experience.

This observational study was approved by the ethical review committee of the local medical association (Institutional Review Board of the Saarland Medical Council, Germany; ID: 232/12) and was conducted in accordance with the Declaration of Helsinki and the Professional Code for Physicians of the local Medical Council. All patients were informed of the purpose of the study by the examiner and signed a form of consent.

Clinical procedure of documentation of clinical findings and sampling

Patients were asked about their case history, nicotine abuse, diabetes presence or absence, regular use of mouth rinses (at least once per week) and use of antibiotics in the past 12 months and if periodontitis had been diagnosed.
For each patient, a pooled sample was taken at three sterile paper points from the peri-implant sulcus around infected or healthy implants. (Fig. 1). Any existing signs of inflammation (pocket suppuration) were documented. In patients with multiple implants, samples were collected for each implant separately, but the data were later pooled to prevent an unbalanced dataset.
If severe peri-implantitis was diagnosed, explantation was performed under local anaesthesia using articaine with epinephrine 1:100.000 (Citocartin Sopira^®, Heraeus Kulzer GmbH, Hanau, Germany).

All patients were informed in advance about the clinical procedure to be performed. In addition, the patients were informed about obtaining paper point samples for a future microbial analysis in connection with the study. All patients gave their informed consent in writing.

For microbial analysis of the peri-implant samples, bacterial DNA from the paper point of a total of 53 implants from 48 patients was obtained and analysed for taxonomic composition by sequencing the V1-V2 variable regions of the 16S rRNA gene. Sequencing was performed at Max Rubner-Institut, Karlsruhe, Germany.

The paper points were applied for 20 seconds according to the established clinical routine for sample collection of peri-implant pathogens (26).

Evaluation at the patient level

Samples were obtained from each implant using three sterile paper points and then pooled. The data acquisition and microbiological evaluation occurred at the implant level. For some patients, multiple implants were analysed. In these cases, the microbial data were later pooled for statistical analysis, effectively creating an average microbial composition of the implants obtained from the same patient. The evaluation was therefore carried out at the patient level.

Extraction of microbial DNA

Sterile paper points were used to collect biofilm samples of the peri-implant sulcus for microbiota analysis, as described previously (26). Implants were also collected after explantation, and the microbiota present was analysed for control purposes. Briefly, biofilm microbes were resuspended in nuclease-free water by subjecting the paper points (or implants) to a combination of shaking and sonication. The suspension was centrifuged, and the pellets were stored at -80°C and then used for DNA extraction using commercial extraction protocols for genomic DNA (QIAamp Mini Kit, Qiagen, Hilden, Germany). The collected pellets were treated with 180 µL lysozyme solution (20 mg/mL, Sigma-Aldrich, Taufkirchen, Germany; 20 mM Tris-HCl, pH 8.0, 2 mM EDTA, 1.2% Triton) under shaking at 37°C for 2 h, 15 min, followed by proteinase K digestion (20 µL proteinase K and 200 µL buffer AL) for 1 h, 15 m under shaking at 56°C. Finally, the DNA was eluted with 100 µL PCR-clean water, and the concentration was quantified using NanoDrop equipment (PEQLAB, Erlangen, Germany). One empty extraction without any sample material was used as a control for background DNA contamination (contamination control).

Illumina sequencing of amplicons targeting the 16S rRNA gene

Amplicons of the V1-V2 region of the bacterial 16S rRNA gene were prepared as published elsewhere (27). Briefly, the genomic sequence of the 16S rRNA gene was amplified with primers derived from previously described primers 27F and 338R (28, 29) and contained sequences compatible with Illumina sequencing platforms and a 6-nt barcode sequence. The resulting DNA was used as the template for a second PCR with primers designed to introduce full-length Illumina adapter sequences, including Illumina 6-nt index sequences, to enable high-level multiplexing. To control for potential DNA contamination, an additional sample was amplified without template DNA (contamination control sample). Libraries were pooled and subjected to 250 nt paired-end sequencing using an Illumina MiSeq machine. For data analysis, the obtained raw reads were first demultiplexed based on barcode sequences, and primer sequences were removed using a Perl script, thereby removing reads that did not contain the primer sequence. Furthermore, the reads were processed using dada2 version 1.16.0 (30). Dada2 determines amplified sequence variants (ASVs) by removing sequencing errors, effectively denoising the data, which tends to produce more accurate results than the commonly used operational taxonomic units (OTUs) (31). Briefly, sequence reads were trimmed to a length of 200 nt (forward read) and 150 nt (reverse read), and 5 nt of the left end was additionally trimmed. Reads with ambiguous base calls and an expected error rate larger than 2 were discarded, and ASVs were inferred using the pseudo-pooling algorithm. Chimaeric sequences were removed, and ASVs were classified using the RDP naïve Bayesian classifier algorithm, as implemented in dada2, against the Silva database v138 (32). ASVs were assigned species names using the assignSpecies function of dada2, if sequence data allowed exact matching. After preprocessing, the data amounted to between 1,305 and 205,073 sequences per sample.

Statistical analysis of microbiome data

Further analysis of the microbiome data was performed in R version 3.6.1 (36) using the packages phyloseq (37), vegan (38) and ggplot2 (39). The contamination control sample yielded 397 reads, and 13 ASVs represented by more than 5 sequences in the contamination control were removed from all samples. The alpha diversity of the sampled biofilms was estimated using the inverse Simpson index, as calculated by the phyloseq function estimate_richness. Principal coordinate analyses (PCoAs) were performed using the ordinate function with Jenson-Shannon divergence (JSD) as the distance metric. Differences in single variables between groups of samples were assessed using the Kruskal-Wallis rank sum test (R function kruskal.test) for multiple groups and the Wilcoxon rank sum test (R function pairwise.wilcox.test) for pairwise comparisons of groups with Benjamini-Hochberg adjustment to control for the false discovery rate in multiple comparisons (40). Differences in the microbiota composition (beta diversity) were examined by performing permutational multivariate analysis of variance (PERMANOVA) with the JSD matrix using the adonis function of the R package vegan (38) and pairwise Adonis by Martinez Arbizu & Monteux (41). Differential species abundance was calculated by merging ASVs that were classified as belonging to the same species (or genus for those ASVs that could not be classified at the species level) and by comparing two groups with the R package ALDEx2 v1.20.0 (42) using the functions aldex.ttest and aldex.effect. Species were defined as differentially abundant if the p-value of the Welch test was less than 0.05.

The study was conducted in accordance with the STROBE guidelines.

Data availability

The raw sequence data were uploaded to the European Nucleotide Archive (ENA, https://www.ebi.ac.uk/ena) with accession No. PRJEB41299.

Composition of the peri-implant microbiota

Across all samples, 552 taxa (amplified sequence variants, ASVs) that occurred with an abundance of at least 1% in one or more samples were detected. Most of the identified taxa belong to the six phyla Actinobacteriota, Bacteroidota, Firmicutes, Fusobacteriota, Proteobacteria and Spirochaetota (nomenclature according to version 138 of the Silva database). All samples could be analysed and were included in the study.

The microbial composition of the peri-implant biofilm samples showed substantial variation between individual patients, and most ASVs were not found in all patients, even when comparing only those of the same group (Fig. 2). Representatives of bacterial genera found in most samples (though at variable abundance) included Streptococcus and Fusobacterium. Most of the Streptococcus ASVs could not be classified at the species level, with a few exceptions, such as S. sanguinis and S. canis. Some taxa were found rarely or not at all in most samples but were found at high abundances in single samples. This included extreme cases, where ASVs of a single genus (Rothia, Aggregatibacter and Gemella) comprised more than 50% of the sequences. Despite the large variability, comparison of the composition across all samples revealed a pattern that partially matched the different groups of patients (Figs. 2 and 3). Most notably, most of the patients who lost their implants (groups E and L) could be distinguished from their respective control groups (Fig. 3b, c). While the samples of group L clustered together on the right side of the plot (with two exceptions), the samples of group E showed large differences. Some of the latter clustered with the L samples and some with the control samples. The difference in the composition of the different groups was statistically significant, either when all groups were compared with each other or only the implant loss groups with their respective controls (p < 10^-5, PERMANOVA test on beta diversity). Patients with implant loss showed increased average abundances of Fusobacterium and Porphyromonas (Fig. 2b), and the most abundant species of these genera, F. nucleatum and P. gingivalis, contributed largely to the placement of the respective samples in the PCoA plot (Fig. 3). The two healthy control groups, on the other hand, harboured comparatively high abundances of bacteria of the genera Streptococcus, Neisseria, Rothia and Veillonella (Fig. 2b, Fig. 4). Analysis of beta diversity compares microbial compositions at a global level. We also compared data at the species level to provide further insight into the differential abundance of certain bacteria under different conditions. Comparison of the late-loss groups with each other and their respective control groups resulted in a limited list of differentially abundant species (Table 1). Five species were found to be significantly higher in group L than in group E, namely, Desulfobulbus sp., Pseudoramibacter alactolyticus, Treponema sp., Fretibacterium sp. and a species of Anaerovoracaceae, but the effects were quite small. Comparing groups E and L with their controls, 3 and 10 species were found to be differentially abundant, respectively. Some bacteria were found to be differentially abundant in two of these comparisons, Desulfobulbus sp., P. alactolyticus and Fretibacterium, all of which were found to be higher in late-loss group L than in the control group CL. In addition, Streptococcus sanguinis was reduced in E compared to the control CE, whereas Streptococcus sp. was reduced in L compared to CL.

A number of parameters were collected for each patient, some of which might impact microbial composition (Suppl. Table 1). First, we evaluated whether the parameters were evenly distributed among the groups. The parameters implant region, diabetes, antibiosis (within 12 to 2 months prior to sampling), and regular mouthwash use showed no significant deviation from an even distribution (p > 0.05, Fisher’s exact test). The type of implant system used showed some bias, however, with the majority of implants being of type “Astra”, which were overrepresented in groups E and CE (p = 0.040). Additionally, smoking was unevenly distributed (p = 0.048), with smokers being overrepresented in groups E and L (77 and 78%, respectively) compared to CE and CL (35 & 40 %). Another important factor is periodontitis, which was diagnosed if found anywhere in a patient and not only close to the implant that was analysed. Patients with implant loss had a higher incidence of periodontitis (E: 64 %, L: 71 %) than the control groups (CE: 29 %, CL: 20 %). The impact of these potential confounders on the microbial data was examined by adding each confounder as an independent variable to the PERMANOVA test for beta diversity between groups. Neither antibiosis nor periodontitis had a significant effect on beta diversity, and the differences between groups were still significant in extended models. Similarly, diabetes, implant system smoking and regular mouthwash use had no significant effect.

Biofilm samples were also analysed directly from implants after explantation and compared to the corresponding paper point samples to validate the results. Differences in the microbial composition were highly similar between implant and paper point samples. Although multivariate analysis of the composition (PERMANOVA) did show a significant difference between implants and paper points (R² = 0.021, p = 0.016), it was small compared to the difference between the groups (R² = 0.107, p < 10^-4).

In recent decades, several risk factors for the failure of dental implants have been reported – most of them being clinical aspects of implant treatment, medications or health conditions. With regard to clinical risks for bone loss around implants, surgical skills or mistakes regarding the surgical protocol, such as dull burrs, high pressure while drilling or high insertion torque, different healing protocols or excess cement can cause problems (43-46). On the other hand, medications, such as bisphosphonates or antidepressants, can be risk factors for infections or implant loss (47-49). Additionally, certain conditions, such as poor bone quality, periodontitis, bruxism, overload or mechanical complications and smoking habits, can have an impact on the osseointegration and long-term success of dental implants; however, controlled diabetes, osteoporosis or the age of the patient do not seem to be risk factors (50-57).

Among reasons for implant failure, bacterial infections have been reported to be a major risk factor for early and late loss of dental implants (7, 55, 58, 59). Most studies to date focus on microbial findings in peri-implantitis, whereas there are no data for early loss. Differences between research groups may be explained by different test systems and by the steady development of microbial tests. In general, microbial findings in peri-implantitis resemble those of periodontitis, with some notable differences (17, 19, 20, 59). Although certain known periodontal pathogens may also be associated with the aetiology of peri-implantitis, apparently, there are many differences between these two clinical conditions involving distinct microorganisms (20).

The results obtained in this study present a very heterogeneous picture of the microbes found around failed dental implants. A huge majority of this variation can likely be attributed to differences between patients, with no association with any clinically relevant parameters. Overall, the statistical analysis proved difficult due to the huge underlying variation and limited sample size of the study groups, resulting in only a few bacterial groups showing differential between-group abundances that were statistically significant. Nevertheless, both early- and late-loss samples showed some similarity and, on average, could be distinguished from their respective control groups (Fig. 3). Therefore, the group of late-loss samples (L) formed a more compact group, with the exception of two samples, whereas early-loss samples appeared to be more heterogeneous.

According to our results, pathogens such as Tannerella forsythia and Treponema sp. were frequently found around implants in the group of late-loss implants, which had to be removed because of severe bone loss after a primarily uneventful osseointegration. Additionally, Porphyromonas gingivalis and Fusobacterium nucleatum were found in both early and late loss, though the latter involved several sequence variants (ASVs) all classified as the same species, only some of which contributed to the communities in the late-loss samples. The statistical analysis also revealed increased abundances of Desulfobulbus sp., Fretibacterium sp. and Pseudoramibacter alactolyticus, which were previously found in peri-implant sites (60, 61)

For early loss of implants, the microbial results were very heterogeneous, and only a subset of these cases were similar to the major proportion of late-loss samples, while a large fraction resembled the healthy controls. Some of the samples in this group were each dominated by a single genus, notably Fusobacterium, Aggregatibacter and Gemella, which accounted for more than 50 % of the microbial sequences found in these samples (Fig. 4). Such samples might partially explain the observation that implants lost early had lower microbial diversity (Fig. 2a) than implants lost late. This heterogeneity between patients points towards multiple potential risk factors contributing to early implant losses, which might include i) characteristic microbial communities, ii) dysbiosis involving other oral bacteria and iii) other, non-microbial causes such as surgical mistakes or poor bone quality.

Antibiosis is included in many implantological protocols. This approach has an impact on the oral microbiome (62). Therefore, patients who had taken antibiotics within the last two months prior to explantation were excluded from this study. Nonetheless, approximately half of the patients used antibiosis within 12 to 2 months prior to sampling. We therefore tested the effect of antibiosis on the microbial composition but found no differences between patients with and without antibiosis.

In the present study, regular use of mouthwash did not have any significant impact on the peri-implant microbiome. However, it is known that mouthwash does affect the peri-implant microbiome (22), and heterogeneity within the groups and the small number of cases may have influenced the present results. There were a number of additional potential confounding variables that might have affected the microbial composition in the patients examined, as follows: use of diabetes, smoking, differences in the implant system and, especially, periodontitis. The distribution of smokers and implant systems was indeed biased between the different groups. Smokers were overrepresented in both groups with implant loss, but there were fewer in the control groups; no significant difference in microbial composition was observed. This might hint towards an effect of smoking on the outcome, which is independent of microbial colonization. The implant systems were biased, with the System Astra being overrepresented in the “younger” implants of groups E and CE; this might be because these implants were placed by the same surgeon. Astra was the main implant system in this clinic at the time. For most other systems, only a few were implanted; thus, we cannot draw conclusions on this factor, and the microbial composition did not differ between the systems. A potentially important factor for the occurrence of inflammation and the source of oral pathogens is the diagnosis of periodontitis. The observation that patients with implant loss were more likely than those of the control groups to also be diagnosed with periodontitis indeed suggests that the disease may be an important contributing factor. Notably, we nevertheless found no significant difference in the composition of implant-associated biofilm in patients with periodontitis, even though some periodontal pathogens were found to be abundant in patients with implant loss. In a mixed model with periodontitis and the experimental group as independent factors, the microbial composition was different between the groups but not between patients with and without periodontitis. Thus, although periodontitis seems to be a potential factor for implant failure, the composition of the implant-associated biofilm appears to be independently associated.

Because of the sample size, the data presented should be interpreted as a first indication, especially regarding the heterogeneous findings for early implant loss. It can also not be distinguished whether bacterial colonization is the reason for the loss of the implant or if it represents secondary infection after osseointegration fails or is lost. This underlines the importance of further studies including a larger number of implants and also assessing the different potential risks of early losses.

In summary, samples from late-loss implants formed a more homogeneous group in terms of microbial composition, whereas early-loss samples appeared to be more heterogeneous. In general, implants lost early also showed lower microbial diversity than implants lost late. Nevertheless, the microbial results were not indicative of the causes of early and late losses.

ASVs - amplified sequence variants

DNA - deoxyribonucleic acid

F. nucleatum - Fusobacterium nucleatum

HbA_1c - hemoglobin A1c

JSD - Jenson-Shannon divergence

OTUs - operational taxonomic units

PCR - Polymerase chain reaction

PERMANOVA - performing permutational multivariate analysis of variance

P. gingivalis - Porphyromonas gingivalis

PCoA - Principal coordinates analysis

S.- Streptococcus

16S rRNA - Sequences of bacterial ribosomal RNA

Ethics approval and consent to participate: The Ethics Committee of the Saarland Medical Council reviewed and approved the proposed study (Ref. No.: ID: 232/12). The participants were informed about the study and that participation in it is voluntary. In addition, information was provided about the positive ethical vote on this survey. The declaration of consent was obtained in writing. Participants who did not want to participate in the study declined to give informed consent.
Consent for publication: The authors agree to the publication under the guidelines of BMC.
Availability of data and materials: We provide the data as supplementary files.
Competing interests: The authors declare no conflicts of interest.
Funding: The study was not funded.
Author contributions:

Statement listing the contributions made by each of the authors

KM: Design, data analysis/interpretation, article drafting, approval of the submitted version of the manuscript.
MSM: Design, data collection, Data analysis/interpretation, article drafting, approval of the submitted version of the manuscript.
SD: Data analysis/interpretation, article drafting, approval of the submitted version of the manuscript.
PC: Data analysis/interpretation, article drafting, approval of the submitted version of the manuscript.
DA: Design, statistics, data analysis/interpretation, article drafting, approval of the submitted version of the manuscript.

All authors have read and approved the manuscript.

Acknowledgements: The authors would like to thank Melanie Huch, Luisa Martinez and Lilia Rudolf for excellent technical assistance. Furthermore, we thank Prof. Matthias Hannig for his support in planning the study and revising the manuscript.

Moraschini V, Poubel LA, Ferreira VF, Barboza Edos S. Evaluation of survival and success rates of dental implants reported in longitudinal studies with a follow-up period of at least 10 years: a systematic review. Int J Oral Maxillofac Surg. 2015;44(3):377-88.
Chambrone L, Chambrone D, Lima LA, Chambrone LA. Predictors of tooth loss during long-term periodontal maintenance: a systematic review of observational studies. J Clin Periodontol. 2010;37(7):675-84.
Corbella S, Taschieri S, Samaranayake L, Tsesis I, Nemcovsky C, Del Fabbro M. Implant treatment choice after extraction of a vertically fractured tooth. A proposal for a clinical classification of bony defects based on a systematic review of literature. Clin Oral Implants Res. 2014;25(8):946-56.
Friedman PK, Kaufman LB, Karpas SL. Oral health disparity in older adults: dental decay and tooth loss. Dental clinics of North America. 2014;58(4):757-70.
Hamood E. The evaluation of success and failure of endodontic treatments. Australian endodontic journal : the journal of the Australian Society of Endodontology Inc. 2001;27(2):80-4.
Rinke S, Ohl S, Ziebolz D, Lange K, Eickholz P. Prevalence of periimplant disease in partially edentulous patients: a practice-based cross-sectional study. Clin Oral Implants Res. 2011;22(8):826-33.
Sakka S, Baroudi K, Nassani MZ. Factors associated with early and late failure of dental implants. Journal of investigative and clinical dentistry. 2012;3(4):258-61.
Renvert S, Persson GR. Periodontitis as a potential risk factor for peri-implantitis. J Clin Periodontol. 2009;36 Suppl 10:9-14.
Berglundh T, Armitage G, Araujo MG, Avila-Ortiz G, Blanco J, Camargo PM, et al. Peri-implant diseases and conditions: Consensus report of workgroup 4 of the 2017 World Workshop on the Classification of Periodontal and Peri-Implant Diseases and Conditions. J Periodontol. 2018;89 Suppl 1:S313-s8.
Venza I, Visalli M, Cucinotta M, De Grazia G, Teti D, Venza M. Proinflammatory gene expression at chronic periodontitis and peri-implantitis sites in patients with or without type 2 diabetes. J Periodontol. 2010;81(1):99-108.
Rodriguez-Argueta OF, Figueiredo R, Valmaseda-Castellon E, Gay-Escoda C. Postoperative complications in smoking patients treated with implants: a retrospective study. J Oral Maxillofac Surg. 2011;69(8):2152-7.
Costerton JW, Stewart PS, Greenberg EP. Bacterial biofilms: a common cause of persistent infections. Science. 1999;284(5418):1318-22.
Fux CA, Costerton JW, Stewart PS, Stoodley P. Survival strategies of infectious biofilms. Trends Microbiol. 2005;13(1):34-40.
Heitz-Mayfield LJ, Lang NP. Comparative biology of chronic and aggressive periodontitis vs. peri-implantitis. Periodontol 2000. 2010;53:167-81.
Davey ME, Costerton JW. Molecular genetics analyses of biofilm formation in oral isolates. Periodontol 2000. 2006;42:13-26.
Muyzer G, de Waal EC, Uitterlinden AG. Profiling of complex microbial populations by denaturing gradient gel electrophoresis analysis of polymerase chain reaction-amplified genes coding for 16S rRNA. Appl Environ Microbiol. 1993;59(3):695-700.
Charalampakis G, Belibasakis GN. Microbiome of peri-implant infections: lessons from conventional, molecular and metagenomic analyses. Virulence. 2015;6(3):183-7.
Renvert S, Lindahl C, Renvert H, Persson GR. Clinical and microbiological analysis of subjects treated with Branemark or AstraTech implants: a 7-year follow-up study. Clin Oral Implants Res. 2008;19(4):342-7.
Persson GR, Renvert S. Cluster of bacteria associated with peri-implantitis. Clin Implant Dent Relat Res. 2014;16(6):783-93.
Faveri M, Figueiredo LC, Shibli JA, Perez-Chaparro PJ, Feres M. Microbiological diversity of peri-implantitis biofilms. Adv Exp Med Biol. 2015;830:85-96.
Perez-Chaparro PJ, Duarte PM, Shibli JA, Montenegro S, Lacerda Heluy S, Figueiredo LC, et al. The Current Weight of Evidence of the Microbiologic Profile Associated With Peri-Implantitis: A Systematic Review. J Periodontol. 2016;87(11):1295-304.
Daubert DM, Weinstein BF. Biofilm as a risk factor in implant treatment. Periodontol 2000. 2019;81(1):29-40.
Daubert D, Pozhitkov A, McLean J, Kotsakis G. Titanium as a modifier of the peri-implant microbiome structure. Clin Implant Dent Relat Res. 2018;20(6):945-53.
Jacobi-Gresser E, Huesker K, Schütt S. Genetic and immunological markers predict titanium implant failure: a retrospective study. Int J Oral Maxillofac Surg. 2013;42(4):537-43.
Rakic M, Grusovin MG, Canullo L. The Microbiologic Profile Associated with Peri-Implantitis in Humans: A Systematic Review. Int J Oral Maxillofac Implants. 2016;31(2):359-68.
Korsch M, Marten SM, Walther W, Vital M, Pieper DH, Dotsch A. Impact of dental cement on the peri-implant biofilm-microbial comparison of two different cements in an in vivo observational study. Clin Implant Dent Relat Res. 2018;20(5):806-13.
Camarinha-Silva A, Jauregui R, Chaves-Moreno D, Oxley AP, Schaumburg F, Becker K, et al. Comparing the anterior nare bacterial community of two discrete human populations using Illumina amplicon sequencing. Environ Microbiol. 2013.
Lane DJ. 16S/23S rRNA sequencing. Nucleic acid techniques. In: Stackebrandt E, Goodfellow M, editors. Bacterial systematics. Chichester, UK: Wiley & Sons; 1991. p. 115-75.
Etchebehere C, Tiedje J. Presence of two different active nirS nitrite reductase genes in a denitrifying Thauera sp. from a high-nitrate-removal-rate reactor. Appl Environ Microbiol. 2005;71(9):5642-5.
Callahan BJ, McMurdie PJ, Rosen MJ, Han AW, Johnson AJ, Holmes SP. DADA2: High-resolution sample inference from Illumina amplicon data. Nat Methods. 2016;13(7):581-3.
Callahan BJ, McMurdie PJ, Holmes SP. Exact sequence variants should replace operational taxonomic units in marker-gene data analysis. ISME J. 2017;11(12):2639-43.
Quast C, Pruesse E, Yilmaz P, Gerken J, Schweer T, Yarza P, et al. The SILVA ribosomal RNA gene database project: improved data processing and web-based tools. Nucleic Acids Res. 2013;41(Database issue):D590-6.
Schloss PD, Westcott SL, Ryabin T, Hall JR, Hartmann M, Hollister EB, et al. Introducing mothur: open-source, platform-independent, community-supported software for describing and comparing microbial communities. Appl Environ Microbiol. 2009;75(23):7537-41.
Kozich JJ, Westcott SL, Baxter NT, Highlander SK, Schloss PD. Development of a dual-index sequencing strategy and curation pipeline for analyzing amplicon sequence data on the MiSeq Illumina sequencing platform. Appl Environ Microbiol. 2013;79(17):5112-20.
Rognes T, Flouri T, Nichols B, Quince C, Mahe F. VSEARCH: a versatile open source tool for metagenomics. PeerJ. 2016;4:e2584.
R Core Team. R: A Language and Environment for Statistical Computing: R Foundation for Statistical Computing; 2019 [Available from: https://www.R-project.org.
McMurdie PJ, Holmes S. phyloseq: an R package for reproducible interactive analysis and graphics of microbiome census data. PLoS One. 2013;8(4):e61217.
Oksanen J, Blanchet FG, Friendly M, Kindt R, Legendre P, McGlinn D, et al. vegan: Community Ecology Package (version 2.5-1) 2018 [Available from: https://CRAN.R-project.org/package=vegan.
Wickham H. ggplot2: Elegant Graphics for Data Analysis: Spinger-Verlag New York; 2009.
Benjamini Y, Hochberg Y. Controlling the False Discovery Rate: a Practical and Powerful Approach to Multiple Testing. J R Statist Soc B. 1995;57(1):289-300.
Martinez Arbizu P. pairwiseAdonis: Pairwise multilevel comparison using adonis. R package version 0.0.1. 2017 [Available from: https://github.com/pmartinezarbizu/pairwiseAdonis.
Fernandes AD, Reid JNS, Macklaim JM, McMurrough TA, Edgell DR, Gloor GB. Unifying the analysis of high-throughput sequencing datasets: characterizing RNA-seq, 16S rRNA gene sequencing and selective growth experiments by compositional data analysis. Microbiome. 2014;2:15.
Al-Sabbagh M, Bhavsar I. Key local and surgical factors related to implant failure. Dental clinics of North America. 2015;59(1):1-23.
Chrcanovic BR, Kisch J, Albrektsson T, Wennerberg A. Impact of Different Surgeons on Dental Implant Failure. Int J Prosthodont. 2017;30(5):445-54.
Troiano G, Lo Russo L, Canullo L, Ciavarella D, Lo Muzio L, Laino L. Early and late implant failure of submerged versus non-submerged implant healing: A systematic review, meta-analysis and trial sequential analysis. J Clin Periodontol. 2018;45(5):613-23.
Korsch M, Obst U, Walther W. Cement-associated peri-implantitis: a retrospective clinical observational study of fixed implant-supported restorations using a methacrylate cement. Clin Oral Implants Res. 2014;25(7):797-802.
Gupta B, Acharya A, Pelekos G, Gopalakrishnan D, Kolokythas A. Selective serotonin reuptake inhibitors and dental implant failure-A significant concern in elders? Gerodontology. 2017;34(4):505-7.
de-Freitas NR, Lima LB, de-Moura MB, Veloso-Guedes CC, Simamoto-Junior PC, de-Magalhaes D. Bisphosphonate treatment and dental implants: A systematic review. Med Oral Patol Oral Cir Bucal. 2016;21(5):e644-51.
Chrcanovic BR, Kisch J, Albrektsson T, Wennerberg A. Factors Influencing Early Dental Implant Failures. J Dent Res. 2016;95(9):995-1002.
Chrcanovic BR, Albrektsson T, Wennerberg A. Smoking and dental implants: A systematic review and meta-analysis. J Dent. 2015;43(5):487-98.
Chrcanovic BR, Albrektsson T, Wennerberg A. Bone Quality and Quantity and Dental Implant Failure: A Systematic Review and Meta-analysis. Int J Prosthodont. 2017;30(3):219-37.
Moraschini V, Barboza ES, Peixoto GA. The impact of diabetes on dental implant failure: a systematic review and meta-analysis. Int J Oral Maxillofac Surg. 2016;45(10):1237-45.
de Medeiros F, Kudo GAH, Leme BG, Saraiva PP, Verri FR, Honorio HM, et al. Dental implants in patients with osteoporosis: a systematic review with meta-analysis. Int J Oral Maxillofac Surg. 2018;47(4):480-91.
Zhou Y, Gao J, Luo L, Wang Y. Does Bruxism Contribute to Dental Implant Failure? A Systematic Review and Meta-Analysis. Clin Implant Dent Relat Res. 2016;18(2):410-20.
Nguyen-Hieu T, Borghetti A, Aboudharam G. Peri-implantitis: from diagnosis to therapeutics. Journal of investigative and clinical dentistry. 2012;3(2):79-94.
Sendyk DI, Rovai ES, Pannuti CM, Deboni MC, Sendyk WR, Wennerberg A. Dental implant loss in older versus younger patients: a systematic review and meta-analysis of prospective studies. J Oral Rehabil. 2017;44(3):229-36.
Ting M, Craig J, Balkin BE, Suzuki JB. Peri-implantitis: A Comprehensive Overview of Systematic Reviews. J Oral Implantol. 2018;44(3):225-47.
Koutsonikos A. Implants: success and failure--a literature review. Annals of the Royal Australasian College of Dental Surgeons. 1998;14:75-80.
Pokrowiecki R, Mielczarek A, Zareba T, Tyski S. Oral microbiome and peri-implant diseases: where are we now? Ther Clin Risk Manag. 2017;13:1529-42.
Sanz-Martin I, Doolittle-Hall J, Teles RP, Patel M, Belibasakis GN, Hammerle CHF, et al. Exploring the microbiome of healthy and diseased peri-implant sites using Illumina sequencing. J Clin Periodontol. 2017;44(12):1274-84.
Koyanagi T, Sakamoto M, Takeuchi Y, Ohkuma M, Izumi Y. Analysis of microbiota associated with peri-implantitis using 16S rRNA gene clone library. J Oral Microbiol. 2010;2.
Ferrer M, Mendez-Garcia C, Rojo D, Barbas C, Moya A. Antibiotic use and microbiome function. Biochem Pharmacol. 2017;134:114-26.

Table 1: Differential abundance of species between different groups
Species name	FDR	abund L	abund E	effect
Pseudomonas sp.	0.0070	1.86	6.50	-1.18
Anaerovoracaceae uncl.	0.0240	8.82	4.07	0.94
Desulfobulbus sp.	0.0142	8.82	-0.44	1.15
Pseudoramibacter alactolyticus	0.0079	7.23	-0.37	1.21
Treponema sp.	0.0151	10.48	1.62	1.04
Pantoea sp.	0.0064	2.69	7.89	-1.13
Fretibacterium sp.	0.0170	9.21	2.00	1.02
Shewanella sp.	0.0078	2.75	7.55	-1.09

Species name	FDR	abund L	abund CL	effect
Rothia sp.	0.0187	2.41	10.56	-1.06
Streptococcus sp.	0.0063	9.20	12.48	-1.52
Desulfobulbus sp.	0.0209	8.79	-1.74	1.73
Pseudoramibacter alactolyticus	0.0256	7.18	-1.40	2.10
Fretibacterium sp.	0.0555	9.18	-0.55	1.55
Haemophilus parainfluenzae	0.0015	2.36	8.62	-1.65
Haemophilus sp.	0.0035	2.63	9.32	-1.52
Dialister pneumosintes	0.0333	4.93	-1.98	1.43
Capnocytophaga sp.	0.0930	2.14	7.00	-0.94
Granulicatella sp.	0.0299	-0.09	5.87	-1.12

Species name	FDR	abund E	abund CE	effect
Streptococcus sanguinis	0.0061	3.69	9.99	-1.40
Lautropia mirabilis	0.0539	1.60	7.90	-1.00
Actinomyces naeslundii	0.0961	0.43	6.26	-0.84

Legend
abund X: relative abundance of species in group X (clr transformed)
FDR: False discovery rate (Benjamini Hochberg-corrected p value, Welch-test)
effect: ratio of the median between-group differences and within-group differences uncl.: indicates that this taxon was not classified at a lower rank

Microbiological findings in early and late implant loss: an observational clinical cohort study

Status:

Journal Publication

Version 2

Abstract

Figures

Background

Methods

Results

Composition of the peri-implant microbiota

Discussion

Conclusions

Abbreviations

Declarations

References

Table

Supplementary Files

Status:

Journal Publication

Version 2