Clinical and Histological Outcomes of Maxillary Sinus Floor Augmentation With Synthetic Bone Substitutes for Dental Implant Treatment: A Meta-Analysis

Endosseous dental implant placement is a challenging procedure when there is insufficient bone volume in the posterior upper jaw.¹  Maxillary sinus floor augmentation (MSFA) using lateral or crestal approaches has been proposed for increasing bone volume and placing dental implants with a high level of success.² 

Although autologous bone (AB) is considered the gold standard in grafting material for bone reconstructive surgery, it does have setbacks that frequently lead to other bone replacement alternatives being selected.^3,4  A wide variety of bone graft materials are currently being used to augment vertical and horizontal bone gain in preparation for dental implant placement, including AB, xenografts, allografts, synthetic bone substitutes (SBS), or even employing a graftless technique in cases of MSFA.

SBS are used because of their high biocompatibility and osteoconductive potential.⁴  These materials can be bioactive glass, calcium sulfate, polymers, or metals, and they do not require a donor site, thereby reducing morbidity and operation procedure time.⁵  Their osteoconductive properties play an essential role in blood clot stabilization in the initial healing stages⁶  by offering better conditions during the graft maturation period.⁷  Unfortunately, SBS are one of the most expensive grafts among alternative graft materials.⁸ 

SBS represent a large group of inorganic biomaterials with different chemical compositions and physical properties, which are used as a 3-dimensional structural scaffold for bone regeneration and cell growth.⁹  Osseointegration with the natural bone and a mature and sturdy structure with the newly formed bone have been observed, making them widely used in medical fields such as oral and maxillofacial surgery.^10,11 

The ideal SBS should have chemical, physical, and mechanical properties similar to the tissue where they will be implanted.⁸  They therefore need to be biocompatible, osteoconductive, and have a similar biodegradation rate to the original bone as well as have an adequate porous structure (pore size, pore volume, and interconnectivity) for cell growth and differentiation.¹¹ 

Equivalent rates of implant success have been reported when using AB grafts or SBS such as tricalcium phosphate.¹  However, there is still no consensus on what material should be used in MSFA procedures.³  Thus, the objective of this meta-analysis was to analyze the relevant data from randomized clinical trials to assess the clinical and histological results of MSFA with SBS to determine the most reliable use in dental implant procedures.

This systematic review was performed by following the PRISMA guidelines (Preferred Reporting Items of Systematic Review and Meta-Analyses).¹² Table 1 depicts the PICOS question.

The study included randomized controlled clinical trials (RCTs) with at least 1 year of follow-up and published in English since 1990. We selected articles that compare clinical dental implant parameters of maxillary sinus floor augmentation grafted with synthetic bone substitute versus MSFA grafted with other bone substitutes in healthy patients. We recorded the following clinical variables: implant success rate (%), implant failure (%), peri-implant marginal bone loss (mm), bone gain (mm), and number of surgical complications. In addition, newly formed bone (%) and residual graft material (%) were extracted from the selected studies as histological variables. Studies with an insufficient description of the outcomes were excluded.

Two independent reviewers (J.T.-S. and A.R.-F.) carried out an electronic search in the PubMed/MEDLINE, Cochrane Library, and Scopus databases between October and November 2019. The search strategy was (“Sinus Floor Augmentation” [Mesh] OR “sinus lift” OR “maxillary sinus floor augmentation” OR “maxillary sinus augmentation” OR “MSFA” OR “sinus floor elevation”) AND (“Synthetic Bone Substitutes [Mesh] OR “replacement material bone” OR “alloplast” OR “synthetic graft” OR “beta-tricalcium phosphate” OR “β-TCP” OR “biphasic calcium phosphate” OR “calcium phosphate” OR “hydroxyapatite” OR “HA” OR “nanohydroxyapatite” OR “bioactive glass” OR “HTR”) AND (Dental Implants [Mesh] OR “dental implant”). Additionally, a manual search was conducted in the following journals for publications in the last 10 years: Oral Surgery, Oral Medicine, Oral Pathology, and Oral Radiology, Clinical Oral Implants Research, Implant Dentistry, Journal of Oral Implantology, International Journal of Oral and Maxillofacial Implants, Clinical Implant Dentistry, and Related Research.

The study selection process was carried out by two independent reviewers (J.T.-S. and A.R.-F.). First, the studies were selected, assessing the abstract when the title was relevant. Full-text analysis was conducted on those with apparent relevance or when the abstract was unavailable. Finally, only studies that fulfilled the inclusion criteria were included in the present review. Cohen's kappa coefficient (κ) was calculated to measure the reviewers' level of agreement.

After the data extraction process, a qualitative synthesis was displayed in tables. The following items were retrieved by two reviewers (J.T.-S. and A.R.-F.) from the included articles: total number of patients, implants, and MSFA, implant success rate, implant failure, peri-implant marginal bone loss (mm), vertical bone gain (mm), newly formed bone (%), residual graft material (%), graft healing time (months), follow-up time (years), and intra- and postoperative complications.

Once the final articles were selected, the risk of bias of each paper was done by means of the Cochrane Handbook for Systematic Reviews of Interventions.¹³  Two reviewers (J.T.-S. and A.R.-F.) assessed independently the risk of bias, evaluating (1) random sequence generation, (2) allocation concealment, (3) patient blinding, (4) outcome blinding, (5) incomplete outcome data addressed, and (6) selective reporting. Authors were contacted for clarification of missing or unclear information when necessary.

A meta-analysis was conducted with RevMan 5.3 (Review Manager version 5.3; The Cochrane Collaboration, Copenhagen, Denmark). A quantitative synthesis was performed using odds ratio (OR) with 95% CI for dichotomous outcomes to estimate the effect of the intervention. For continuous outcomes we used mean differences (MDs) and standard deviation to summarize data for each group. In split-mouth designs, the implants used in each pair were the unit of analysis, and in parallel groups the unit of analysis were the patients.¹⁴ 

Pairwise meta-analysis (PMA) was carried out when studies compared the same outcome measures. The random effect model was selected because methodological and clinical heterogeneity was expected across the included studies that compared the same pairs of interventions.¹⁵  Data from split-mouth studies was combined with data from parallel group trials using the generic inverse variance method.¹⁶  Statistical heterogeneity was estimated by I²  analyses.¹⁷  An I²  value of >50% was interpreted as significant heterogeneity. Statistical significance was defined as P < .05 for all analyses.

The initial electronic database search yielded 276 articles and one additional article was included after a manual search of the reference lists of relevant articles. After removing duplicate articles and assessing the title and abstract, a total of 21 studies were selected for full-text analysis. Reviewer agreement was 92% with a k index of 0.84 (almost complete agreement).

Twelve publications were excluded after applying study inclusion criteria due to insufficient data,^18–21  no comparison of bone materials,^22–25  and design.^26–29  Finally, 10 RCTs were selected for qualitative synthesis^30–39  and 4 for quantitative synthesis^30,37–39  (Table 2).

Figure 1 shows a flowchart diagram of the screening process.

Only one article was considered to have low risk of bias.³⁶  The remaining studies were classified as having unclear risk of bias mainly owing to selection and performance bias.^{30,32,33,37,38}  This may, therefore, influence the results we obtained, which should be interpreted with caution. Figure 2 summarizes the quality of the RCTs included.

The 10 studies selected included 616 implants in 288 patients, 15 of which dropped out in the follow-up period that ranged from 1 to 5 years.^31,37,38  MSFA procedures were carried out using a lateral approach and a resorbable collagen membrane to cover the bone defect and prevent the invasion of soft tissue into the grafted sites in 7 of the articles.^{30,33,35–39} 

None of the studies revealed significant differences between the groups in terms of implant success,^{31–33,35–39}  implant failure rates,^31–39  and peri-implant marginal bone loss.^31,37,39 

In one trial, autogenous bone (AB) mixed with deproteinized bovine bone (DBB) showed higher bone gain than micro and macroporous biphasic calcium phosphate combined with fibrin sealant (MBCP-FS) (MD: −1.40; 95% CI: −1.9 to −0.9 mm; P < .001).³¹ 

With regard to new bone formation, 3 trials reported significant differences in favor of AB at the end of the follow-up period.^31,34,35  However, one article showed greater bone formation among patients treated with SBS compared to xenografts.³⁸  Of the 10 studies that compared SBS and AB with xenografts, 3 reported greater residual graft material in the experimental groups,^31,33,36  though one showed statistically significantly higher graft resorption in the control group when using xenografts.³² 

A total of 17 implant failures were described in the included articles. Ten (58.82%) occurred in the experimental group, showing 4 implant failures in the BCP group, 3 in the MBCP-FS group, and another 3 were in the HA group. There were 7 failures in the non-synthetic treated sites (41.18%).^{30–33,37,39} 

In 360 MSFA, the only intraoperative complication that occurred in the included studies that were included were Schneiderian membrane perforations.^31,36,38  Additionally, 28 postoperative adverse events were reported in this review,^31–33  with intraoperative bleeding, gingival recession at the implant site, or swelling at the surgical area being the most common complications.

Table 3 shows the main results as well as the details of the studies included in this review.

We conducted a meta-analysis with only 4 articles comparing BCP and DBB groups. We were not able to draw other possible comparisons between SBS and other bone grafts. From the included studies, one combined a parallel group and split-mouth design³⁸  and the rest had a split-mouth design.^30,37,39  The 4 articles included 110 patients, 56 of which were treated with BCP and the remaining cases were treated with DBB.^30,37–39 

Three studies provided information on implant success outcome, involving a global sample size of 193 implants in 100 patients.^37–39  Results of a meta-analysis showed a mean success of 91.6% and 95.6% for BCP and DBB groups, respectively. Quantitative analysis did not reveal significant differences in implant success (OR: 0.48; 95% CI: 0.11 to 2.09; P = .33; I² = 0%) when comparing both bone grafts (Table 4).

All the studies included in this review provided data on implant failure outcome, which included a global sample size of 225 implants in 110 patients.^30,37–39  Quantitative analysis did not reveal any differences in implant failure (OR: 1.22; 95% CI: 0.28 to 5.38; P = .79; I² = 0%) when BCP and DBB groups were compared (Table 4).

Two studies provided data on peri-implant marginal bone loss outcome, involving a global sample size of 94 implants in 40 patients.^37,39  No statistically significant differences in the peri-implant marginal bone loss outcome were found (MD: 0.21 mm; 95% CI: −0.17 to 0.59; P = .28; I² = 0%) (Table 4).

Two studies provided information on newly formed bone outcome, involving a global sample size of 78 patients.^37,38  The pooled data showed that there were no statistically significant differences between the BCP and DBB groups that were observed with regard to newly formed bone outcome (MD: 1.73 mm; 95% CI: −4.33 to 7.79; P = .58; I² = 44%) (Table 4).

Two studies provided information on residual graft material outcome, involving a global sample size of 78 patients.^37,38  The mean effect revealed that residual graft material was higher in patients treated with DBB than those that received BCP (MD: −4.80 mm; 95% CI: −9.35 to −0.26; P = .04; I² = 0%). However, none of the articles included showed significant differences (Table 4).

The objective of this systematic review was to test the hypothesis of whether SBS are the most appropriate alternative to AB or DBB for obtaining the best clinical and histological results in MSFA procedures. Although we initially wanted to compare SBS with the other graft alternatives, the available data made that impossible, and we were only able to compare BCP with DBB. Finally, a meta-analysis of 4 randomized clinical trials involving patients who had undergone MSFA was carried out. With regard to clinical outcomes, no differences between BCP and DBB groups were found in implant success, implant failure, and peri-implant marginal bone loss. Furthermore, although the DBB group showed a greater amount of residual bone graft, we did not find any differences in the variable of new bone formation.

Some of the study's limitations were related to the different implant systems and grafting materials used in the selected studies, which may influence the clinical and histological outcomes that were evaluated. However, we must consider that the overall heterogeneity of the study did not reveal significant value across the selected studies. In addition, our results are mostly based on papers with unclear risk of bias, so they should be interpreted with caution.

Peri-implant MBL may be affected by physical or biomechanical factors, including graft properties.⁴⁰  However, the overall peri-implant MBL was similar in the analyzed groups. It is worth mentioning that the use of different implant manufacturers across the selected studies are confounding factors that may influence our results.

By comparison, Kim and Yun⁴¹  reported a higher risk of implant infection when SBS were used instead of AB, especially if they were not covered by a membrane. We did not see these findings, as there were 10 implant failures in the SBS group without any significant differences in the studies that compared the SBS and the AB groups.

Histomorphometrical analyses are important methods for evaluating the healing process of the bone substitute grafts²¹  that help describe their osteoconductive ability,^27,34  maturation graft time, and new bone formation.⁴²  Depending on the type of the graft, healing time may vary with a period ranging between 4 to 9 months.⁴³  Though some studies documented a direct relationship between new vital bone formation and graft maturation time,⁴²  further evidence and research are needed to evaluate this effect, and longer healing times are required in sinuses grafted with SBS.

Along these lines, Iezzi et al²¹  reported the importance of striking a balance between graft resorption and new bone formation. However, the optimum period of time in which complete graft resorption occurs and new bone grows in the pristine bone remains unspecified. New bone formation begins between the synthetic particles rather than on their surface in the premature healing process,⁴⁴  resulting in weaker adhesion and contact with the vital bone in contrast with the xenografts.⁴⁵  Several trials included in this review showed that the AB group was more effective in creating new bone than the different SBS,^31,34,35  which can be explained by the osteogenic potential of AB grafts.

Bone resorption behavior of the bone graft has been studied in our review. Histologically, SBS showed less new bone formation and presented a greater percentage of residual graft particles in contrast to sites treated with AB.^31,36,46  This is strongly reflected in the Danesh-Sani et al³⁶  trial, which showed an average of 32% (BCP) and 4.8% (AB) of residual graft particles, but 28.2% and 36.8% of new bone formation, respectively. Conversely, some studies confirmed that SBS have greater resorption compared with xenografts, with the bone grafts having the lowest degradation rate.^25,37  Along these lines, our results seem to support these findings since the graft resorption rates were significantly higher in the BCP groups compared to the DBB groups.^37,38 

BCP can be used as an alternative to DBB in MSFA procedures and exhibit similar clinical results in implant success, implant failure, and peri-implant marginal bone loss after a 1-year follow-up. Though DBB shows a greater amount of residual graft than BCP histologically, no differences between these materials existed in newly formed bone.

Well-designed, long-term randomized clinical trials are required to clarify the question of which type of bone graft is the safest and most effective.

Abbreviations

AB:

autologous bone

BCP:

biphasic calcium phosphate

CI:

confidence interval

DBB:

deproteinized bovine bone

HA:

hydroxyapatite

MBCP-FS:

macroporous biphasic calcium phosphate combined with fibrin sealant

MBL:

marginal bone loss

MD:

mean difference

MSFA:

maxillary sinus floor augmentation

OR:

odds ratio

PMA:

Pairwise meta-analysis

RCT:

randomized clinical trial

SBS:

synthetic bone substitutes

This study has been performed by the research group, Odontological and Maxillofacial Pathology and Therapeutic, of Bellvitge Biomedical Research Institute (IDIBELL). The authors deny any conflicts of interest related to this study.