Extraction of a natural tooth inevitably results in significant three-dimensional alveolar bone resorption, particularly within the first 6 months [1,2]. As a consequence, alveolar ridge resorption can complicate the placement of implants into prosthetically driven positions. Therefore, to limit the post extraction dimensional changes, alveolar ridge preservation therapies, which intend to preserve the ridge volume in the envelope existing at the time of extraction, have been proposed [3].

The results of previous systematic reviews indicate that alveolar ridge preservation is effective at minimizing bone reduction [4-6]. In particular, alveolar ridge preservation performed immediately after tooth extraction were shown to result in significantly less vertical and horizontal contraction of the alveolar crest compared to the spontaneous healing [5]. Additionally, the beneficial effect of alveolar ridge preservation was more pronounced in the prevention of horizontal bone resorption, followed by the prevention of resorption in the vertical mid-buccal and vertical mid-lingual bone [4,6].

With regard to clinical outcomes of implants placed in grafted sites, the implant survival rates following lateral bone augmentation or sinus-floor elevation with the lateral approach were reported to be high and comparable to the ones placed into native bone [7,8]. Likewise, based on previous clinical studies, implants placed simultaneously with guided bone regeneration procedures exhibited clinical performance (with respect to survival rates, marginal-bone-height, and peri-implant soft tissue parameters) similar to that of implants in non-grafted sites [9-11]. Moreover, lateral bone augmentation procedures were found to be associated with peri-implant tissue stability after short-term (1 to 3 year) and long-term (> 3 year) follow-ups [12].

Until now, however, there has been limited evidence regarding the clinical outcomes of implants inserted following ridge preservation. Hence, the aim of this systematic review is to assess the existing evidence regarding the clinical outcomes of implants placed into previously grafted extraction sockets.

Protocol and registration

The methods of the analysis, inclusion and exclusion criteria were specified in advance and documented in a protocol. The review was registered in PROSPERO, an international prospective register of systematic reviews.

The review protocol was developed according to the PRISMA (Preferred Reporting Items for Systematic Review and Meta-Analyses) statement [13].

Focus question

The following question was developed according to the population, intervention, comparison, and outcome (PICO) study design (Table 1):

What are the clinical and radiographic treatment outcomes of the dental implants placed into the grafted sockets?

Information sources

The electronic databases MEDLINE (PubMed) was searched for relevant articles published until 1st April 2019.

In MEDLINE (PubMed), the search was limited to humans, and English language. Also, a filter concerning "Article Type" was applied: controlled clinical trial, randomized controlled trial, evaluation studies, clinical study, clinical trial, comparative study, multicenter study, observational study, twin study, validation studies.

In addition, a hand search was performed including reference lists of all full-text articles and the following scientific journals: "Clinical Oral Implants Research", "Clinical Implant Dentistry and Related Research", "European Journal of Oral Implantology", "Implant Dentistry", "International Journal of Oral & Maxillofacial Implants", "International Journal of Periodontics and Restorative Dentistry", "Journal of Clinical Periodontology", "Journal of Oral Implantology", "International Journal of Oral and Maxillofacial Surgery", "Journal of Periodontology", "Journal of Prosthetic Dentistry", "Open Dentistry Journal", "Journal of Implants and Advanced Clinical Dentistry".

Search

The following search terms were used: (dental Implants [Mesh] OR implant) AND (bone regeneration [Mesh] OR socket preservation [Mesh] OR socket graft [Mesh] Or ridge preservation [Text word] OR augmentation [Text word]).

Selection of studies

During the first literature-selection stage, the titles and abstracts of all identified studies were screened for eligibility by two independent reviewers (AR and AC).

Inclusion and exclusion criteria

The following inclusion criteria were applied:

At the second stage, the full texts of potentially eligible articles were reviewed and evaluated according to the following exclusion criteria:

Differences between reviewers were resolved by discussion and consensus. The level of inter-examiner agreement for the first- and second literature-selection stages was expressed by Cohen’s kappa-score.

Data extraction and data items

From the selected articles fulfilling the inclusion criteria, the following data were retrieved into pre-defined tables:

Quality assessment

The Cochrane Collaboration’s tool for assessing risk of bias was used in the case of controlled clinical trials [15] (Table 5). In six categories (random sequence generation, allocation concealment, blinding, incomplete outcome data, selective reporting, other potential risk of bias) a rating of low, unclear or high risk of bias was performed.

Data synthesis

A meta-analysis integrates the quantitative findings from separate but similar studies and provides a numerical estimate of the overall effect of interest. All meta-analyses were performed on randomized controlled clinical trials that reported the clinical outcomes of non-surgical periodontitis treatment utilizing different adjunctive antiseptics.

Thus, each study provided estimates of outcome measures. The goal was to obtain global estimates of these measures and to test whether they differed significantly. Global estimates of a proportion can be obtained by simply pooling together the data from each study. However, a test for significance cannot be applied to such pooled data, as these studies are heterogeneous with respect to study population and treatment protocol. Therefore, individual trials were pooled, and the weighted mean differences for the MBL changes, together with their 95% confidence intervals (CIs), were calculated. Under the fixed-effects model, it is assumed that all studies come from a general population and that the size is not significantly different among the different trials. This assumption was tested by the heterogeneity test using the Cochran Q statistics. We considered that in our case the random-effects model (the Der Simonian and Liard method) [16] was more appropriate to use since it took into account both the random variation within the studies and the variation among different studies. Later findings indicated the fixed-effects model might be invalid. Indeed, the random-effects model tended to give a more conservative estimate (i.e, with a wider CI), but the results from the 2 models usually agreed well.

Search results

The initial electronic search resulted in the identification of 115 titles. Following the evaluation of titles and abstracts, 99 publications were excluded (Cohens kappa = 0.927). The remaining 16 full-text articles were evaluated. The reasons for exclusion were as follows: (i) studies included < 10 patients in the Test group [17-19], did not assess the outcomes of dental implants [20,21], follow-up period < 1 year [22-26]. Finally, 8 articles were included into the analysis (Figure 1).

Characteristics of the included studies

The included studies are described in Tables 1 to 3. Two of the studies involved the same patient sample that was evaluated at different follow-up periods [27,28], therefore, a total of 7 original clinical investigations were included. All of them were designed as randomized controlled trials with a follow-up period ranging from 1 to 4 years. Control group participants in 3 of the studies had implants placed into naturally healed sites [29-31]. In the remaining 4 investigations Control group participants were treated with immediately placed implants [27,32-34] or were inserted 6 weeks post extraction [34]. Additionally, in the latter 4 investigations, immediate implants were inserted either into grafted sockets [27] or the gap between the implant and the bone was filled with a bone-fill material [32-34].

This systematic review pooled data collected from 512 patients (243 men, 270 women). All investigations included systemically healthy patients with a mean age range of 42.7 to 55.8 years. Smokers were included in 4 of the studies [30,32-34]. Smoking habit was an exclusion criteria in one investigation [27] and patient smoking status was not reported in two of the studies [29,31]. In 5 of the 7 studies, untreated periodontitis [27,32-34] or a history of periodontitis [29] was an exclusion criteria. Patients’ periodontal status was not reported in two studies [30,31] (Table 2).

Socket grafting in all seven studies involved guided bone regeneration concept with the use of bone substitute material covered with a barrier membrane (Table 3). As a bone substitute, either xenogenous (six studies) or alloplastic (one study) bone-filler particles were used. As a barrier membrane, either collagen membranes (in five studies), porcine derma (one study) or a pericardium membrane (one study) were used. In all but one study [29], all patients received prophylactic antibiotics preoperatively and continuously 4 to 6 days postoperatively (Table 4).

Of the total of 520 implants included, 232 were inserted into the previously grafted sockets (3 to 7 months following the grafting) (test), fifty-seven were inserted into the naturally healed sites, and 231 were placed immediately (or 6 week post extraction) after tooth extraction (controls). The diameter of the implants ranged from 3.3 to 5.5 mm, and one investigation included only implants with a diameter of 7 mm [27]. In 5 of the studies, patients were enrolled into either a personalized or regular maintenance program with 3- to 6-month intervals [27,30,32,34].

The assessment of risk of bias in the included randomized trials is presented in Table 5. A low risk of bias was noted in 5 studies according to 4 (3 studies), 5 (2 studies) or all 6 domains (1 study) [27,30,31,33,34]. The remaining 2 studies were judged to have an unclear risk of bias based on 2 (‘allocation concealment’ and ‘blinding’) domains [29,32].

The reported clinical outcomes of the implants

Implant survival, which was defined as a primary outcome, ranged from 95 to 100% for the Test group implants and from 92 to 100% for the implants in the Control groups (Table 4).

The mean MBL changes were reposted in all 7 investigations included. Two of them revealed no significant difference between the implants placed into the naturally healed sites or the previously grafted sockets [29,30]. In contrast, a single study reported higher MBL loss for the implants placed into the naturally healed sites versus grafted sockets [31]. In five investigations where immediate implants were the Control group, three of them found significantly higher MBL resorption for the implants placed into previously grafted sockets, compared to the immediately inserted ones [32-34]. Contradictory findings were reported by Tallarico et al. [27], where immediately placed implants experienced significantly higher rates of the MBL loss.

A pink aesthetic score (PES) was reported in five of the included studies [27,31-34]. In two of them, no difference was noted between the implants placed into the grafted sockets or immediately placed ones [32,33], whereas one study reported significantly higher PES scores for the immediately inserted implants [34]. Conversely, two investigations found superior soft tissue measurements for the implants placed into the healed, previously grafted sockets, compared to the immediately placed ones [27] and implants placed into the naturally healed sites [31]. Additionally, it has to be pointed out that none of the included studies reported the clinical parameters (i.e., BOP, PD) or the occurrence of peri-implant diseases (i.e., peri-implant mucositis or peri-implantitis). Moreover, three of the investigations revealed that additional bone grafting procedures were more often necessary around the implants placed into the naturally healed sites, compared to the previously grafted post extraction sockets [29-31]. Two of these studies found this difference to be statistically significant [29,30].

Meta-analysis

A meta-analysis was conducted to assess MBL changes in the following groups:

The results showed significantly higher marginal bone loss for the implants in previously non-grafted healed sites than for the ones placed in the grafted sockets (weighted mean difference (WMD) = -1.961, 95% confidence interval [CI] [-3.793 to -0.13]; degrees of freedom [df] = 3; heterogeneity test [Q] = 54.36; P < 0.0001) (Figure 2).

The comparison between implants in the grafted sockets vs. immediate implant placement (or 6 weeks post extraction) demonstrated no significant difference between the two implant groups (WMD = 0.194, 95% CI [-0.342 to 0.731]; df = 4; Q = 28.29; P < 0.0001) (Figure 3).

The aim of the present systematic review was to investigate the clinical performance of the implants placed in the previously grafted sockets. The included clinical studies presented methodological differences in terms of the variables assessed to measure the outcomes, and the Control groups (immediate implant placement vs. implants in the healed non-grafted sites).

Based on the included 7 randomized clinical trials, the implants’ survival rates ranged between 95% and 100% after 1 to 4 years of implant placement, with no significant differences between implants placed into previously augmented sockets and those inserted into the spontaneously healed sites. These findings corroborate the survival rates noted for the implants placed into the laterally augmented alveolar ridge sites (78.2 to 100%) [35] and for implants inserted into the laterally augmented sinuses (88.6 to 100%) [8]. Furthermore, the implant-survival rates following socket grafting were within the range of those previously reported for implants placed in pristine sites (93.2 to 100%) [7].

The present meta-analysis points toward significantly higher MBL loss for the implants in previously non-grafted healed sites than for implants placed in grafted sockets. Furthermore, the 3 studies included in the meta-analysis revealed that additional bone-grafting procedures were often required during implant placement at the sites without previous socket grafting [29-31]. These findings are comparable with the results of the previous meta-analysis, in which implant placement without bone augmentation resulted in significantly higher radiographic marginal bone loss compared to at the sites where dehiscence-type defects were augmented using a xenograft bone substitute and a collagen membrane [36]. However, out of three studies included in the present quantitative analysis, significant differences in terms on the MBL changes were reported in the investigation with the longest follow-up period (4 years [31] versus 1 [29] and 3 years [30]). Therefore, the follow-up period might be a critical factor in the treatment outcomes.

Further comparison of the MBL changes between implants in the grafted sockets vs. immediate implant placement demonstrated no significant difference between the two approaches. Similar results were obtained in a previous systematic review, in which implant placement into fresh extraction sockets versus delayed implants into healed sockets resulted in comparable MBL changes [37]. However, among the 4 investigations included in the present meta-analysis, an inconsistency regarding the treatment outcomes was noted. Specifically, while Tallarico et al. [27] found significantly higher MBL loss for the immediately inserted implants over those placed into the healed and previously grafted sockets, the remaining investigations reported significantly higher MBL for the implants in the latter group (i.e., with delayed implants in the previously grafted sockets) [32-34].

The soft-tissue changes around the implants were assessed by PES score in 5 of the included studies [27,31-34]. Again, inconsistency existed among the results of the included studies. In particular, while 2 studies found superior PES scores for the implants placed into the healed, previously grafted sockets [27] and implants placed into the naturally healed sites [31] as compared to the immediately placed ones, Esposito et al. [34] contrarily presented significantly better soft-tissue outcomes for the immediately inserted implants. The latter study corroborates the results of the previous 3-year clinical investigations, which found superior soft-tissue outcomes (PES scores) for the immediately placed and immediately restored implants, as compared to the conventionally restored implants installed in to the healed sites [38]. The authors noticed that superior aesthetic outcomes could be expected in young patients (≤ 30 years) with implants in central incisor/cuspid areas and in the presence of bone recontouring [38].

Contradictory findings exist regarding the timing of implant placement and soft-tissue treatment outcomes.

A recent prospective clinical investigation reported no significant difference in PES values between the immediate and conventional implant placement after 1- and 8 years of follow-up [39]. This observation is supported by the 2 included studies in the present analysis that revealed comparable soft-tissue outcomes for the immediate and delayed implant placement [32,33].

Recently, a meta-analysis has found the comparable prevalence of peri-implant mucositis and peri-implantitis as well as implant loss for dental implants placed in either native or augmented sites (including vertical and/or horizontal ridge augmentation) at both the patient and implant levels [40]. In the present analysis, however, none of the studies reported the clinical parameters or the occurrence of biological implant complications, which prevents comparison with the previous investigations.

In terms of survival rates, placing dental implants in previously grafted sockets is a predictable treatment option. Differences could be found for the peri-implant marginal-bone-level changes when compared with the implants placed into the non-grafted sites. Other objective quantitative methods must be considered in future studies to assess the peri-implant tissue changes that occur over time using different treatment approaches, in terms of the timing of grafting procedures and of implant placement.