Survival of tooth-implant connections: A systematic review and meta-analysis

Abstract

Implant-supported prostheses have considerable biomechanical advantages in partially edentulous patients when compared to other prosthetic options. Given the steady drop in the frequency of patients reporting with complete edentulism, it is not unusual to see situations where teeth and implants can be splinted to provide support for fixed prostheses. A tooth implant prosthesis differs majorly from an implant-supported prosthesis in terms of force dissipation and design. The aim of this systematic review was to compare the survival rates of tooth-implant-supported prostheses with fully implant-supported and fully tooth-supported prostheses. Using the appropriate search terms, PubMed, Google Scholar, and other indexed journals were used to search the English-language literature. According to the review protocols and the PICOS inclusion criteria, the pertinent studies were chosen. The screening of appropriate studies, evaluation of study quality, and data extraction were carried out independently by two reviewers. The pooling of survival data by prostheses failure, implant failure, and marginal bone loss was used in the meta-analysis. The cumulative data of all included studies indicated that tooth-implant-supported prostheses showed a 5-year survival rate of 77%–84% and a 10-year survival rate of 72%. The pooled risk ratio for prostheses failure and implant failure was 0.99 and 1.76, respectively. These results were not statistically significant (P > 0.05). The pooled standard mean difference for marginal bone loss was 0.59, and the results were statistically significant (P < 0.05). A tooth-implant-supported fixed partial denture (FPD) has a similar survival rate when compared to implant-supported FPD or T-FPD.

INTRODUCTION

The occurrence of complete edentulism among patients has shown a consistent decline in recent years, coinciding with a growing demand for fixed dental prosthetic rehabilitation. The introduction of endosseous implants proved to be a game changer in the world of prosthodontics. Historically and conventionally, fixed partial denture (FPD) has either derived support completely from two teeth or two implants. However, it is not uncommon to come across cases where implants can potentially be splinted to natural teeth to provide support for an FPD. This is commonly seen in Kennedy Class I and Kennedy Class II situations. When the distribution, condition, or number of the natural teeth that is still present makes it difficult to restore the mouth with conventional fixed prostheses, teeth can be joined to implants to achieve this.

A combination of implant and tooth support for FPD is suitable, according to Belser et al.[1] In 1986, Ericsson et al. stated that an endosseous implant may be utilized as an extra abutment to natural teeth for FPD.[2] This implant can be used as an abutment for standard single-tooth implant prostheses, or it can be coupled with a neighboring natural tooth to support a fixed partial denture.[3]

To withstand the biomechanical forces applied to the tooth/restoration at the crestal bone region, a natural tooth’s support structures are better constructed. The presence of enamel and dentine, the periodontal membrane, the nerve and blood vessel complex, the biomechanical design of the tooth root and material, and the inherent characteristics of the surrounding bone reduce the risk of occlusal overload to the natural tooth system. However, this is not true for dental implants. The periodontal ligament acts as a shock-absorbing cushion in the natural tooth, while in implants, there is no such interface present as there is functional bony ankylosis between the implant and the surrounding bony architecture. The junctional epithelium and the connective tissue complex are far more fragile around the dental implant. Differences in the biological width, vascularity, and pressure sensitivity are also noted between implants and the natural tooth.[4] There is an absence of a periodontal ligament around an osseointegrated implant and whenever there is splinting of two structurally different components, one with a periodontal ligament, i.e. the tooth, and one without the periodontal ligament, i.e. the dental implant, there will be an unequal force dissipation. This unequal force dissipation can be attributed to the lack of any sort of micromovement around a dental implant. On application of 0.1 N of force, a natural tooth with a healthy periodontal ligament exhibits mobility of 50–200 μ, whereas this displacement is less than 10 μ for an implant.[1]

It was suggested that nonrigid connectors (NRCs) be used in their place because the difference in mobility raised concerns about whether it was possible to rigidly attach natural teeth to implants.[5] The implant will support more of the occlusal load if the connection is rigid, with the natural abutment serving as a cantilever. This may result in increased crestal bone loss.[6] Contradictory to the above-mentioned disadvantages of a rigid connector, prostheses that used an NRC reported a higher rate of failure.[2]

The major advantage of tooth-implant-supported fixed partial dentures (TI-FPDs) over implant-supported FPDs (I-FPD) is the increased proprioception provided by natural teeth abutments owing to the presence of periodontal ligament, which gives patients more chewing comfort. “The tactile perception of PDL is 8.8 times greater than implant abutments.”[7] Long-span edentulous gaps, cantilever portions, and teeth with decreased periodontal support and increased mobility can all benefit from the use of TI-FPD.[6] In addition, they could be used in anatomically restricted areas like those with insufficient alveolar bone, close vicinity to the maxillary sinus, and close proximity to the inferior alveolar nerve canal.[8-10] Such anatomic complexities demand for additional surgical procedures which could be contraindicated or refused by the patient for financial and other reasons. Moreover, exodontia, bone augmentation procedures, or additional risks that are associated with implant placement can be avoided with TI-FPDs. Prostheses deriving support from tooth and implant are significantly more economical than superstructures solely supported by implants. The unique proprioception provided by the periodontal ligament can prevent occlusal overload in tooth implant-supported prostheses.

The main objective of this systematic review was to determine the survival rate of an FPD deriving support from both the implant and a natural tooth.

MATERIALS AND METHODS

A comprehensive systematic review and meta-analysis were carried out. This study followed the “Preferred Reporting Items for Systematic Reviews and Meta-Analyses 2020 (PRISMA 2020),[11] the Cochrane Handbook for Systematic Reviews of Interventions, version 5.1.0, and 4^th Edition of the JBI Reviewer’s Manual”[12] and was registered at PROSPERO under registration code CRD42022340274.

Population, intervention, comparison, and outcome

• Population: Studies including edentulous participants more than 18 years’ age group

• Intervention: Studies including edentulous patients in which rehabilitation is done using TI-FPD in the maxillary or mandibular arches

• Comparison: Studies including edentulous patients in which rehabilitation is done using only I-FPD or only fixed tooth-supported prostheses (T-FPD) in the maxillary or mandibular arches

• Outcome: Studies providing information about the survival of both types of prostheses in terms of the number of prostheses failed or survived. Additional outcomes include changes in periodontal clinical parameters such as clinical attachment loss, probing depth, plaque index, gingival index, marginal bone loss, and any other related outcomes.

Inclusion criteria

Randomized controlled trials (RCTs) or quasi-experimental studies, nonrandomized trials, longitudinal studies, studies with full-text articles, studies providing a numeric value or sufficient information to calculate at least one of the outcome measures previously mentioned, only studies published in English, and studies published until January 2023 were included.

Exclusion criteria

Studies where participants did not give informed consent, studies involving removable prostheses, studies with any comparison group other than only implant-supported and only tooth-supported prostheses, review articles, case studies, in vitro, animal, and studies that only provide an abstract were excluded.

Search strategy

Studies were chosen in accordance with the review protocol’s PICOS inclusion criteria. To find studies that might be eligible, two reviewers assessed the titles and abstracts. A third reviewer was approached for any questions.

The primary outcomes measured were the mean survival rate in the intervention and control groups. The PRISMA for conducting a meta-analysis was followed. The electronic data resources consulted for the elaborate search were PubMed, Google Scholar, and Directory of Open Access Journals (DOAJ). Searches were conducted on all articles published through January 2023, regardless of the language of the publication. The following keywords and MeSH terms were entered in the advanced search option with Boolean operators.

A concept table was prepared based on the PICOS criteria of the review question, and the search strategy was formulated according to it [Table 1].

Table 1.

Concept table

Population	Intervention	Comparison	Outcome	Study design
Adults with partial edentulism	Tooth-implant-supported prostheses	Implant-supported prostheses OR Tooth-supported prostheses	Survival Implant failure Marginal bone loss Plaque index Clinical attachment loss Probing depth	RCT Quasi-experimental studies Non-RCT Retrospective studies

RCT: Randomized controlled trial

Search strategy in PubMed

((tooth-implant-[All Fields] AND supported [All Fields]) AND (fixed [All Fields] AND (“prostheses and implants” [MeSH Terms] OR (“prostheses” [All Fields] AND “implants” [All Fields]) OR “prostheses and implants” [All Fields] OR “prostheses” [All Fields]))) AND (“mortality” [Subheading] OR “mortality” [All Fields] OR “survival” [All Fields] OR “survival” [MeSH Terms]).

Entry terms used in Google Scholar

1. Tooth-implant-supported prostheses

2. Survival.

The above-mentioned was the final search history for the databases accessed till January 2023.

Selection of studies

Each study’s title and abstract were examined and evaluated critically by two separate reviewers. The methods used to apply the selection criteria included integrating the search results to eliminate duplicate entries, looking at titles and abstracts to eliminate articles that were obviously irrelevant, retrieving the full texts of articles that might be relevant, grouping and binding multiple articles from the same study, looking at the full texts of the articles to determine how closely the studies complied with the eligibility criteria, and establishing connections with other studies.

Data extraction

Data were independently gathered by two reviewers from the included studies after focusing on the articles from all the databases. Disagreements were once more settled through conversation. Where there were differences, the third reviewer oversaw a consensus discussion to resolve them. A list was formulated based on the data that were extracted. The main items of this list were authors, year and title of study, country, study design, sample size, age group of the participants, gender, radiographic findings, clinical findings, survival rate, outcomes, results, and other items. All included studies were meticulously and accurately analyzed to extract information about the publication and study, participants, settings, interventions, comparators, outcome measures, study design, statistical analysis and results, and all other pertinent information (funding and conflicts of interest). For each of the individual primary outcomes, data extraction was completed and accurately recorded in the Excel sheets.

Critical appraisal of retrieved studies (risk of bias)

For nonrandomized studies, ROBINS-I[13] checklist was used to perform the quality assessment.

Meta-analysis

Depending on the number of studies providing similar outcomes at the same follow-up interval, a meta-analysis was conducted. The number of events causing survival or failure of prostheses was used as effect sizes for the evaluation of cumulative risk ratios or hazard ratios. The pooled risk ratio was calculated using the fixed model for prostheses failure and marginal bone loss, while the random-effects model was used for implant failure. P < 0.05 was considered statistically significant.

RESULTS

Study selection

The primary electronic database look-up on PubMed/MEDLINE, Cochrane Library, and DOAJ revealed 2040 titles. Out of these, 1511 records were excluded. Two independent reviewers chose 256 pertinent titles from the abstracts; after screening them, 45 were assessed for eligibility. The articles not fitting the inclusion criteria were excluded. The reviewers examined and discussed 16 articles before choosing them for full-text analysis. The chosen studies’ references were manually investigated; however, no additional papers were discovered. After prescreening, application of the inclusion and exclusion criteria, and handling of the PICO questions, 11 studies were included in the quantitative synthesis [Figure 1].

Figure 1.

The Preferred Reporting Items for Systematic Reviews and Meta-Analyses flowchart

Study characteristics

The systematic review included 16 studies. Table 2 depicts the characteristics of 12 studies where the comparison is done between TI-FPD and I-FPD. Table 3 includes two studies where a comparison is done between TI-FPD and tooth-supported prostheses. Table 4 gives the characteristics of three studies where no comparison group is present.

Table 2.

Characteristics of studies comparing tooth-implant-supported and implant-supported prostheses

Study ID	Country	Study design	Sample size	Age (years)	Gender	Follow-up	Outcomes assessed	Survival rate (%)
Gunne, 1991[19]	Sweden	Longitudinal study	23 patients, 69 FPD	Mean: 57.7	Eight males, 15 females	3 years	Implant survival, bridge stability, mobility of teeth, marginal bone levels, plaque, gingivitis, pocket depth, and attachment levels	-
Olsson, 1994[24]	Sweden	Prospective	23 patients, 69 implants	Mean: 58	-	5 years	Implant survival, bridge stability, mobility of teeth, marginal bone levels, plaque, gingivitis, pocket depth, and attachment levels	5 years: 88
Gunne, 1999[23]	Sweden	Prospective	23 patients, 46 FPD, 69 implants	-	-	10 years	Stability, implant stability, BOP, sensory function, and marginal bone level
Bragger, 2000[14]	Switzerland	Prospective	80 patients	Mean: 55.7	53 females, 32 males	4–5 years	Plaque index, gingival index, probing pocket depth, attachment level, and BOP	98.2 survival rate
Hosny, 2000[20]	Belgium	Prospective	18 patients	37–65	All females	14 years	Implant complications, prostheses failure, tooth fracture, and marginal bone level	-
Lindh, 2000[8]	Sweden	Prospective	26 patients, 95 implants	49–84	15 females, 11 males	3, 6, 12, 24 months	Plaque, probing depth, bleeding, marginal bone level, and implant survival	IG: 96 CG: 95
Naert, 2000[9]	Belgium	Prospective	644 implants	20–79	-	15 years	Survival and success	IG: 94.92 CG: 98.41
Romeo, 2004[24]	Italy	Prospective	250 patients, 759 implants	20–67	106 males, 144 females	7 years	Survival and success	90.60
Bragger, 2005[26]	Switzerland	Prospective	89 patients	28–88	34 males, 55 females		Survival and success	I-I type: 54.5 I-T type: 50
Bragger, 2010[27]	Switzerland	Prospective	84 patients, 175 FDPs	-	-	Mean: 11.31 years	Implant survival	-
Rammelsberg, 2012[16]	Germany	Prospective	132 patients, 166 FDP	21–83	-	Mean: 2 years, 4 months	Implant survival	-
Rammelsberg, 2020[17]	Germany	Prospective clinical study	434 FDP	Mean: 60.8	-	Mean: 4.26 years	Failure rate	5 years: 77–84 10 years: 72–77

FDPs: Fixed dental prostheses, FPD: Fixed partial denture, BOP: Bleeding on probing, IG: Implant group, CG: Control group

Table 3.

Characteristics of studies comparing tooth-implant-supported and tooth-supported prostheses

Study ID	Country	Study design	Sample size	Follow-up	Outcomes assessed	Survival rate (%)	Conclusion
Bragger, 2000[14]	Switzerland	prospective	80 patients - three groups	4–5 years	Plaque index, gingival index, probing pocket depth, attachment level, and BOP	98.2 survival rate	Favorable clinical conditions were found at tooth and implant abutments after 4–5 years of function. Loss of FPDs over 4–5 years occurred at a similar rate with mixed, implant-, or tooth-supported reconstructions
Beur, 2015[15]	Germany	Prospective	44 patients, 49 FPD	36 months	Survival, pocket depth	IG: 96.3 CG: 95.5	It can be concluded that tooth-implant-supported FDPs exhibited promising clinical performance. No difference in periodontal parameters compared to tooth-supported FDPs was found

FDPs: Fixed dental prostheses, FPDs: Fixed partial dentures, BOP: Bleeding on probing, IG: Implant group, CG: Control group

Table 4.

Characteristics of studies with only tooth-implant-supported prostheses

Study ID	Country	Study design	Sample size	Age (years)	Gender	Follow-up	Outcomes assessed	Number of prostheses failed	Survival rate (%)	Reasons for failure	Conclusion
Lindh, 2001[28]	Sweden	Retrospective single group	111 patients, 185 implants	45–87	41 females, 40 males	3 years	Implant survival	-	3 years: 95.4	Implantitis, intrusion of the tooth, carious lesions, abutment screw loosening, and fracture of the abutment	The tooth-implant-supported prostheses using the Brånemark System are in the short term, an equally predictable treatment, as the completely implant-supported prostheses concerning implant survival and loss of marginal bone. When combining implants and teeth, a rigid form of connection should be used to prevent tooth intrusion
Nckenig, 2005[21]	Germany	Retrospective single group	83 patients, 84 implants	22–61	-	4.73 months	Survival	2/84	5 years	Not mentioned	Technical complications of implant-supported FPDs are dependent on the different bridge configurations. When using rigid functional connections, similarly favorable values will be achieved as in the case of solely implant-supported FPDs
Chrcanovic, 2020[18]	Sweden	Retrospective single group	85 patients, 96 FPD	-	-	10.5 years	Survival	20	5 years: 90.7 10 years: 84.8 15 years: 69.9 20 years: 66.2	Bruxism, loss of abutments, and cantilevers	Although combined tooth-implant-supported FDPs are an alternative treatment option, this study has found that across 20 years of service, nearly 35% of the prostheses may fail

FDPs: Fixed dental prostheses, FPDs: Fixed partial dentures

Among the included studies, six studies were conducted in Sweden, four in Germany, four in Switzerland, and two in Belgium. From all the studies, 1097 patients in total were included. All the participants were in the 20–88 years’ age group. Implant survival and stability, failure rate, implant complications, marginal bone level, and clinical parameters like plaque index were the outcomes that were measured in these studies.

Survival rate

The cumulative data of all included studies indicated that TI-FPD showed a 5-year survival rate of 77%–84% and a 10-year survival rate of 72%.

Risk of bias

Quality assessment of included studies was done according to the ROBINS-I tool[13] for nonrandomized clinical studies. The risk of bias was assessed across seven domains – confounding bias, selection bias, misclassification bias, bias due to deviation from intended interventions, bias due to missing data, bias due to selective reporting of results, and overall bias [Table 5].

Table 5.

Risk of bias

Study ID	Confounding bias	Selection bias	Misclassification bias	Bias due to deviation from intended interventions	Bias due to missing data	Bias in the measurement of outcomes	Bias due to selective reporting of results	Overall bias
Gunne, 1991[19]	Serious	Low	Low	Low	Low	Low	Low	Moderate
Olsson, 1994[22]	Serious	Serious	Low	No information	Low	Moderate	Low	Serious
Gunne, 1999[23]	Low	Low	Serious	No information	No information	Moderate	No information	Serious
Bragger, 2000[14]	No information	Low	Low	Low	Low	Low	Low	Low
Hosny, 2000[20]	Low	No information	Low	Low	Serious	Low	Low	Moderate
Lindh, 2000[8]	Serious	Serious	Low	No information	Low	Serious	Low	Serious
Lindh, 2001[28]	Low	Low	Low	Low	Low	Low	Low	Low
Naert, 2000[9]	No information	Low	Low	Low	Low	Low	Low	Low
Romeo, 2004[24]	Low	Serious	Low	Serious	Serious	Moderate	Low	Serious
Bragger, 2005[26]	No information	No information	Low	Low	Low	Low	Low	Moderate
Nickenig, 2005[21]	Low	Low	Low	Low	Low	Low	Low	Low
Bragger, 2010[27]	Low	No information	Serious	No information	Low	Moderate	Serious	Serious
Rammelsberg, 2012[16]	Low	Low	Low	Low	Low	Low	Low	Low
Beur, 2015[15]	Low	No information	Low	Low	Serious	Low	Low	Moderate
Chrcanovic, 2020[18]	Low	Low	Low	Low	Low	Low	Low	Low
Rammelsberg, 2020[17]	Low	Low	Low	Low	Low	Low	Low	Low

Out of the included 16 studies, seven studies[8,9,14-18] showed low risk, four studies[18-21] showed moderate risk, whereas five studies[8,16,22-24] showed high risk of bias.

Among the studies with high risk, confounding bias and selection bias were high, which led to high risk among these studies. For studies with moderate risk, bias due to missing data was high, which led to moderate risk among these studies. The risk of bias has been summarized in the traffic light plot [Figure 2] and the summary plot [Figure 3].[25]

Figure 2.

Traffic light plot

Figure 3.

Summary plot

Meta-analysis

Meta-analysis was conducted among studies that provided data on the survival or failure of the prostheses irrespective of the follow-up period. Most of the studies had a mean follow-up of 5 years [Table 6].

Table 6.

Data for meta-analysis

Study ID	Number of prostheses failed		Number of implants failed		Bridge stability		Marginal bone loss
	Tooth implant	Only implant	Tooth implant	Only implant	Tooth implant (%)	Only implant (%)	Tooth implant	Only implant
Gunne, 1991[19]	5 years: 2/69	5 years: 5/69			3 years: 91.3	82.60	2 years: 0.3±0.1	2 years: 0.7±0.2
Gunne, 1999[23]	2/23	4/23	10 years: 2/69	10 years: 6/69	10 years: 85	80	10 years: 0.5±0.6	10 years: 0.7±0.7
Bragger, 2000[14]	1/18	1/40	1/19	1/84	-	-	-	-
Hosny, 2000[20]	No failures				-	-	-	-
Lindh, 2000[8]	2	2			-	-	24 m: −0.09±0.52	24 m: −0.42±0.55
Naert, 2000[9]	10/339	1/305			-	-	-	-
Romeo, 2004[24]	3/31	12/295			-	-	-	-
Bragger, 2005[26]	10 years: 11/22	10 years: 15/33	5/22	1/69	-	-	-	-
Bragger, 2010[27]	10 years: 1/20	10 years: 0/9			-	-	-	-
Rammelsberg, 2012[16]	9/48	34/91			-	-	-	-
Rammelsberg, 2020[17]	8/155	8/213			-	-	-	-

A quantitative assessment was done on the following parameters:

1. Prostheses failure

2. Implant failure

3. Marginal bone loss.

Effect sizes

Effect sizes are quantitative measures of the strength and direction of an intervention’s influence over a given outcome. When studies that evaluate the same outcome do so in different ways, the standardized mean difference is used as a summary statistic. In this case, standardizing the study results to a consistent scale is required before they can be combined.

Prostheses failure

Nine studies[9,14,16,17,19,23,24,26,27] evaluated fixed prostheses failure between TI-S FDP and I-S FDP. The pooled risk ratio was 0.99 [0.71 and 1.38], indicating that the risk of prostheses failure was less with TI-FPD-supported compared to only implant-supported FDPs. This difference was not statistically significant (P > 0.05). The heterogeneity (I²) value was moderate, i.e. I² = 41%. Hence, the fixed-effects model was applied to obtain the pooled risk ratio [Figure 4].

Figure 4.

Meta-analysis of prostheses failure

Implant failure

Four studies[14,22,23,26] evaluated implant failure between TI-S FDPs and I-S FDPs. The pooled risk ratio was 1.76 (0.32 and 9.96), indicating that the risk of implant failure was greater with TI-FPD compared to only I-FPD. These results were not statistically significant (P > 0.05). Heterogeneity was high (I² = 68%); hence, the random-effects model was used [Figure 5].

Figure 5.

Meta-analysis of implant failure

Marginal bone loss

Three studies[8,19,23] evaluated marginal bone loss between TI-S FDPs and I-S FDPs. The pooled standard mean difference was −0.59 (−0.81 and −0.36), indicating that marginal bone loss was greater in the comparison group, i.e. implant-supported as compared to. These results were statistically significant (P < 0.05). Heterogeneity was high (I² = 98%), and a fixed-effects model was applied to obtain the pooled risk ratio [Figure 6].

Figure 6.

Meta-analysis of marginal bone loss

DISCUSSION

Several articles on TI-FPD have been published in recent years. These articles include case–control studies, prospective or retrospective cohort studies, case series, case reports, studies using animals or in vitro models, narrative reviews, and systematic reviews. These articles have produced contrasting findings and opposing viewpoints. To the authors’ knowledge, this is the only systematic review comparing TI-FPD with I-FPD as well as completely tooth-supported prostheses.

The purpose of the systematic review was to determine the success rate of fixed prostheses deriving support from both implants and teeth in patients who needed to replace missing teeth. After a thorough screening of the available literature, 16 studies were selected for the systematic review. Twelve studies had an implant–implant-supported prostheses as the control group,[8,9,14,16,17,19,20,22–24,26,27] three studies were conducted without having any comparison group,[18,21,28] while two studies had tooth–tooth-supported prostheses as the control group.[14,15] All of these studies measured the survival of the TI-FPD. The parameters used to measure the survival of TI-FPD were marginal bone levels, probing depth, attachment levels, bleeding on probing, stability of the prostheses, prosthetic failure, tooth mobility, and fracture. The cumulative data of all included studies indicated that TI-FPD showed a 5-year survival rate of 77%–84% and a 10-year survival rate of 72%–77%.

Gunne et al. conducted a split-mouth study in 1992 comparing the survival of TI-FPD with I-FPD. Twenty-three bridges derived support from both the implant and the tooth, while 23 bridges derived support from both the implant and the tooth. The patients were kept on a follow-up for 3 years.[19] According to Albrektsson, a lack of implant mobility is one of the key factors for success and should be gauged only after the superstructure has been removed.[29] In this study, the implant mobility was assessed without retrieving the prostheses although they concluded that implant survival was the same for both types of bridges.[19] The 5-year survival rate of 88% was reported. Bridge stability was found to be 89% for I-FPD and 91% for TI-FPD. One of the factors that the authors considered to measure the failure of the implant was sensory disturbances. Sensory disturbances commonly relate to nerve injury during the surgical part of the treatment. Nerve injuries are complications, and they seldom indicate the failure of the implant treatment.[23] They concluded that implants and teeth can be connected successfully without having any detrimental effect on either the tooth or the implant.[23]

Hosny et al., in 2000, concluded a 14-year follow-up study evaluating TI-FPD. No implant mobility, fracture of any implant component, and prosthetic complications were observed. They concluded that joining teeth with implants for fixed prostheses had no detrimental effect on long-term survival. This particular study can have a low statistical power owing to its small sample size of just 18 participants and its cross-sectional nature.[20]

A comparative intraindividual split-mouth study was conducted in 2000 on 26 patients with bilaterally missing maxillary posterior teeth. They were treated with two different designs of FPD, namely, I-FPD and TI-FPD. The patients were monitored at intervals of 3, 6, and 24 months. The density of the bone in the affected area is one of the most crucial factors in determining the success of the implant; however, the type of bone in which the implants were placed is not mentioned. TI-FPD and I-FPD, both reported a similar failure rate. Although the survival rate of the control group was high, the follow-up period was too short to draw effective conclusions on actual long-term survival rates.[8]

Brägger et al. concluded a 5- and a 10-year follow-up reporting the survival of TI-FPD. The 5-year survival rate for TI-FPD was 98.2%.[14] This study had 111 patients with 186 implants. The survival rate of TI - FPD was compared with I-FPD as well as T-FPD. There was extreme heterogenicity in the sample size of the three groups which could potentially alter results. TI-FPD had a noticeably higher failure rate. A 50% survival rate of TI-FPD was reported. While TI-FPD experienced more statistically significant technical failures, I-FPD experienced fewer biological complications.[26]

Lindh et al., in 2001, retrospectively studied the survival rate of TI-FPD. One hundred and eleven patients with 185 implants were included in the study. The implant survival rate was found to be 95.4% after 3 years. They also mentioned the use of two different types of connectors to splint teeth and implants; this can potentially introduce some bias in the overall survival rate of implants. Although the survival rate of 95% is excellent, the follow-up duration is still relatively short.[28]

Romeo et al. investigated the long-term success and survival of prostheses that derived support from both the implant and the tooth. They reported a survival rate of 91%.[24]

Although most of the above-mentioned studies reported a greater failure for the TI-FPD, the difference in survival was not statistically significant. Another important parameter that was discussed in some of the studies was marginal bone loss. Three studies[9,20,24] evaluated marginal bone loss between TI-FPD and I-FPD. Marginal bone loss was significantly greater in I-FPD. The majority of these studies concluded that the survival of a TI-FPD is similar to that of an I-FPD or T-FPD, and it can be a viable treatment option whenever indicated.[8,9,14,16,17,19,20,22-24,27] Only one study conclusively indicated a higher percentage of biological and technical complications.[26]

Four studies[14,22,23,26] evaluated implant failure between TI-FPD and I-FPD. The pooled risk ratio suggested that the probability of implant failure was greater with TI-FPD compared to I-FPD. These results were not statistically significant (P > 0.05).

A few of the included articles also discussed the material aspects and the prosthetic outcomes of FPD deriving support from both the implant and the tooth. Nine studies[9,14,16,17,19,23,24,26,27] evaluated fixed prostheses failure between TI-S FDP and I-S FDP. The risk of prostheses failure was less with TI-FPD compared to I-FPD although this difference was not statistically significant.

Nickenig et al. contrasted the technical and biological failures in TI-FPD utilizing two distinct retention techniques. In most cases, the implant abutments were surgically placed in the posterior mandible which frequently has a high bone density when compared to the maxilla. The question is whether the high survival rate reported in this study can be extrapolated to implants placed in the maxilla, which tend to have lower bone densities. They also reported an increased number of complications in screw-retained prostheses than in cement-retained prostheses.[21]

Beuer et al. compared the clinical survival of a three-unit zirconia TI-FPD with T-FPD, and they also discussed the prevalence of biological and technical issues. They were rehabilitated with zirconia and then subsequently veneered with porcelain and cemented with glass ionomer cement. The experimental group only had provisional restorations placed on their natural teeth rather than temporary bridges connecting them to the implants, which was one of the study design’s limitations. It is possible that a small dimensional change can occur between the implant and the tooth during the fabrication of the final restoration. They concluded that zirconia-based FPDs supported by tooth implants and tooth support exhibited comparable clinical performance with similar survival rates.[15]

In a cohort study done in 2021 by Rammelsberg et al., the failure and chipping rates of ceramic and metal-ceramic I-FPDs were assessed over the long term. The likelihood of chipping for prostheses was as follows after 5 years: 3%, 39%, and 18%, respectively, for metal fused to ceramic-fixed dental prostheses with a high noble metal framework.[17]

Chrcanovic et al., in 2020, analyzed technical complications and risk factors for failure for TI-FPD. The prostheses were monitored for 10 years and ranged in span from 2 to 13 prosthetic units. They also compared the implant survival in bruxers and nonbruxers. The major drawback of this study was that the participants who were included in the bruxer group were not given a conclusive diagnosis of bruxism. In addition, there was a lack of information regarding biological aspects such as oral hygiene, bleeding during probing, and probing pocket depth. Comparing bruxers and nonbruxers, bruxers showed a significantly higher rate of prosthetic failure. Eight of the failed prostheses with implant failures were smokers. Thus, these failures cannot be completely attributed to a TI-FPD. The estimated cumulative survival rate of the prostheses was 90.7% (84.6–96.9) and 84.8% (76.8–92.9) at 5 and 10 years, respectively.[18]

A similar systematic review was published in 2020 comparing the survival rate of TI-FPD with I-FPD. In the authors’ knowledge, this is the only systematic review that compares the survival rate of a TI-FPD not only with I-FPD but also with T-FPD.[30] The current review also discusses the marginal bone loss that takes place with a T-I FPD.

There are some limitations to the current systematic review, despite an accurate screening process which might affect the outcomes. The main downside is the complete absence of RCTs. There is also an irregularity in the frequency of published literature. Due to this, it is very challenging to compare the current findings to those that have already been published in the literature. There is also a severe heterogenicity in aims, sample size, implant systems, types of connectors used, design, and material of the prostheses. The follow-up period, which ranges from 6 months to up to 14 years, also varies greatly. Since the majority of these studies were carried out in institutional settings, it is possible that the results cannot be applied to routine dental care.

CONCLUSION

Given the limitations of this systematic review and meta-analysis, the following conclusions can be drawn.

1. The survival rate of TI-FPD and tooth–tooth I-FPD is similar

2. There is no statistically significant difference in prostheses failure with TI-FPD and I-FPD

3. Between TI-FPD and I-FPD, there is no statistically significant difference in implant failure

4. There is an increased marginal bone loss in I-FPD than in TI-FPD.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.