Impact of immediately loaded implant-supported maxillary full-arch dental prostheses: a systematic review

Abstract The immediate loading of implant-assisted fixed prostheses in edentulous maxillae may achieve favorable success rates with reduced treatment time. An evidence summary of clinical trials is key to recommend loading protocols in these cases. Objectives To compare immediately loaded, fully implant-supported complete dentures to early and conventional/delayed loading in the edentulous maxillae of adult patients by a systematic review of controlled clinical trials (CCT). Methodology CCTs reports were identified up to January 17, 2019 from Cochrane Oral Health Group’s Trial register, Cochrane Central Register of controlled trials (CENTRAL), MEDLINE (Ovid), BIOSIS, EMBASE, CINAHL, Web of Science, and DARE. Two independent reviewers screened titles/abstracts and confirmed inclusion using full texts. Data were extracted and quality assessed (Cochrane Risk of Bias tool) independently and in duplicate. Study heterogeneity prevented pooling by meta-analysis. Results Out of 1,052 candidate studies, four CCTs were included. Two trials had patient satisfaction as an outcome: (1) A randomized trial compared immediately and early loaded fixed dentures and found more satisfaction with the first after 12 months; (2) A non-randomized study found better satisfaction with immediate fixed dentures compared to conventional loading after 3 months (no more at 12 months). Regarding implant success and prosthetic complications, three trials did not report significant differences comparing immediate loading to other protocols. Conclusions This review found weak evidence of differences between immediate load and other loading regimens, regarding patient satisfaction and maintenance events/adversities. The potential of immediate loading for favorable results in edentulous maxillae reinforces the need for well-designed RCTs, for solid clinical guidelines. Registration number CRD42018071316 (PROSPERO database).


Introduction
Edentulism poses a major impact on oral and general health, and on quality of life.Edentulous individuals have higher risk of systemic diseases, as pinpointed by the increased mortality rate among the edentulous elderly. 1,2Impaired mastication represents a major consequence of edentulism.Even with goodquality complete dentures, masticatory performance is from 1/5 to 30% of dentate patients. 3,4Besides mastication, conventional denture wearing represents a major psychological and social burden for some patients. 5plant-assisted prostheses can tackle most of the limitations of conventional dentures, regardless of being fixed or removable.Fixed complete prostheses lead to better patient satisfaction in many cases compared to removable alternatives.This is the case when ease of hygiene is not a patient-perceived priority, which is common amongst middle-aged patients. 6Primary indications for fixed prostheses include patients who cannot endure removable dentures or the feeling of being edentulous, strong gag reflex, and history of recurrent sore spots caused by dentures. 7Patients with previous negative denture experience tend to perceive implant-supported fixed prostheses as their own natural teeth, leading to good self-esteem, physical and social well-being. 5spite the focus given to the lower arch, 8 many edentulous patients request conversion of their maxillary dentures to implant-assisted ones.The maxillary arch poses specific challenges, including low-density bone 9 and limiting sinus anatomy. 10rthermore, fixed prostheses are a more intuitive choice for edentulous maxillae, given that patient satisfaction does not seem to improve much with maxillary overdentures. 11e possibility of immediate load after implant insertion can expedite an otherwise time-consuming treatment, i.e. maxillary fixed dentures.Studies have demonstrated high success rates for immediately loaded fixed prostheses in edentulous maxillae, with conventional or zygomatic implants. 10,12Immediate prostheses may be more satisfying for patients than those fabricated by traditional protocols. 13,14 2013, a systematic review of randomized clinical trials (RCT) on immediately loaded implants showed no evidence of different success rates when compared to other loading protocols. 15This review, despite its high quality, did not approach patient-reported outcomes (PRO) (e.g.satisfaction and oral health-related quality of life).Actually, PRO can be considered the main success indicator for prosthodontics. 16Understanding how patients respond to different loading protocols in the edentulous maxilla is essential for developing clinical guidelines.However, there are no systematic reviews considering PROs to understand the effect of these protocols, which would be of primary relevance for clinical recommendations. 17

Methodology
This review was reported according to the PRISMA guidelines (checklist available on Appendix 1). 18A protocol version was published at the PROSPERO database (ID: CRD42018071316). 19econdary outcomes: (1) Specific patient satisfaction items, such as ease in chewing, swallowing, satisfaction with esthetics, and ease of hygiene; Due to the short-term response linked to IL and expected longevity of implant-assisted prostheses, we did not consider a particular timespan.We sought to discuss results for primary outcomes based on shortterm results whenever possible, i.e. within the first three months after loading.

Data extraction and quality assessment
Included studies underwent data extraction and quality assessment by the same authors.We extracted data from trials based on the following characteristics: (1) Study design: time until follow-up, sample size, study setting, sampling criteria, recruitment methods, randomization methods, randomized number, dropouts, withdrawals and losses; (2) Participant: age, gender, general health status (including diabetes mellitus), clinical characteristics (history of periodontitis, maxillary bone volume and density), smoking, drinking habits, other recreational drugs, occlusion during healing phase, previous experience with removable dentures, and attendance to follow-up visits; (3) Intervention and comparators: implants (system, number, type, design, length, positioning, and insertion torque), interim prosthetic design and loading time (if applicable), and definitive prosthetic design and loading time; (4) Outcomes: Assessment method and instrument, baseline and post-treatment scores, as well as time of data collection.
We assessed the quality of included trials by using the Cochrane Risk of Bias tool. 21,22This classifies studies based on six potential sources of bias: If two or more trials reporting the same comparison and outcome were found, we would assess their heterogeneity.In turn, we would synthetize data by meta-analysis if applicable, giving priority to random effect models.We also planned to assess publication bias using a funnel plot, if there were sufficient studies.

Characteristics of included studies
The two trials evaluating patient satisfaction assessed 59 participants, with a single loss (Table 1). 13,14,38Both applied nearly similar inclusion and exclusion criteria on the initial samples of 30 participants/each: edentulous maxillary arches with existing opposing occlusion, not needing augmentation procedures.Lower arches had natural teeth (complete dentition or combined with dental prostheses) or implant-assisted prosthesis.The RCT by Canizzaro, et al. 38 (2008) recruited patients at an Italian private clinic from 2004 to 2005 to compare IL to EL. Follow-up extended to 12 months. 38The non-randomized CCT by Penñarrocha-Oltra, et al. 13,14 (2013, 2014) compared IL to CL. 13,14 Researchers enrolled participants at a

Methodological quality of the trials
All the four trials showed some potential source of bias classified as "high risk".Figure 2  Sequence generation was adequate for Canizzaro, et al. 38 (2008) and Vercruyssen, et al. 41 (2016), whereas only the first was explicit regarding the use of allocation concealment.The other CCTs [Peñarocha-Oltra, et al. 13,14 (2013, 2014) and Tealdo, et al. 39,40 (2011, 2014)] were preference trials; therefore, they were classified at high risk for selection bias-related criteria.
All trials had high risk for performance bias as a limitation -patients cannot be treated blindly, and no study described any approach to prepare prostheses in a way that could mitigate this source of bias.[41] Incomplete outcome data was a minor concern for the four trials.Two trials reported a comprehensive series of outcomes in a way that consistently leads to "low risk" classification for selective reporting. 13,14,38[41] Finally, other potential sources of bias included between-group imbalances regarding: (1) the final prosthesis provided by one of the studies, i.e.
Toronto-type acrylic prostheses, IL: 4 (27%); EL: 9 (60%) participants; 38 and (2) number of implants, i.e.IL received less implants/maxillary denture than the CL group. 39,40One of the preference trials is very unlikely affected by other biases, 13,14 and we could not determine whether sponsorship would influence results of an RCT. 41gure 2-Risk of bias summary for included studies: evaluations on risk of bias concerning each potential source and type of bias.A + signifies that the corresponding approach to minimize bias was probably done (adequately described) for a given study, whereas adiscloses an evident limitation in controlling bias.A question mark underscores that the study provides insufficient description for judging a given approach as adequate or not 2019;27:e20180600 7/15 ABDUNABI A, MORRIS M, NADER SA, SOUZA RF  IL versus CL: Peñarrocha-Oltra, et al. 13,14 (2013, 2014) quantified overall patient satisfaction on a 100mm VAS.The average value after 3 months was 35 mm higher for the IL arm (95%CI: 26 to 44 mm).Such difference recedes after 12-months follow-up (mean difference: 0; 95%CI: -6 to 6 mm).

Discussion
Despite the impact of IL for the management of edentulous maxillae, this review included a small number of CCTs.A recent growing interest in the literature on the subject is evident given the year of the oldest included report (i.e.2008). 38This contrasts with the wide proportion of observational studies on IL found in 2005, 42 thus suggesting a recent shift towards CCTs.
Comparisons between IL and comparators show that patients may be more satisfied when they receive a functional fixed denture, regardless of when.Evidence is minor, but IL was more satisfying than EL in a single RCT after 1 year. 38However, that trial evaluated satisfaction as a secondary outcome and performed a simple assessment.An imbalance in the types of prostheses delivered to the two groups may also have contributed to post-treatment differences.Therefore, as tempting it is to suggest a long-term effect of IL on patient satisfaction, this finding should be interpreted carefully.A trial comparing IL to CL showed similar treatment effect after 3 months; 13,14 this is intuitive, given that participants still had relined conventional dentures in the CL group.Results for satisfaction are similar at 12 months though, suggesting that results may not differ at that point.Patients may get used with existing fixed dentures and provide similar responses after few months.In other words, patients may undergo a response shift and reach similar perception of received prostheses regardless of initial experiences. 43 general, findings suggest that IL is effective compared to EL and CL, although evidence is not enough for solid clinical recommendations.Clinicianreported outcomes show no evident difference in survival rates for implants and prostheses.Failures tend to be quite rare.Other complications show no difference, although a synthesis of the four trials was unviable.Bone loss was not different when IL was compared to other protocols.A trial observed a lower complication rate with IL compared to CL, possibly caused by different prosthetic configurations/n of implants rather than the loading protocol itself.
All studies provided treatment with standard dental implants, thus evidence from CCTs is absent for zygomatic implants.Their potential safety and effectiveness make them a very interesting subject for future trials, as found by observational studies. 10,44,45owever, it is arguable whether results are exactly the same expect for routine patients without research involvement.For instance, potential participants may refrain to participate given potential concerns regarding randomization. 32The inclusion of preference trials may mitigate such issues, by rendering study participants closer to real patients, with freedom to deliberate on which treatment they will receive. 46,47e paucity of studies makes any assumption regarding specific clinical conditions unclear.For example, one cannot infer whether different results are expected because of different occlusal schemes or antagonist arch.The same could not be done for certain adverse conditions that could contra-indicate IL, e.g.severe parafunction, smoking, and high risk of periodontal disease. 48,49 summary, all included studies could be classified as high risk of bias for varying reasons.Amongst design-related issues, the inclusion of preference trials deserves comments, given their important drawback: higher risk of selection bias. 50Those trials cannot implement sequence generation methods able to minimize selection bias.Blinding also was a major limitation, given that blinding the participants and care providers is not possible for the tested comparisons.
In general, studies were careful when reporting the numbers of non-adherent participants.
Given the long-lasting recommendation of trial registration and contemporaneity of included trials, the absence of published protocols was surprising.Trial registration has been a persisting recommendation of guidelines for trial protocols 51 and final reports. 52e of the main limitation of this review is the low number of included studies.A scarcity of RCTs was foreseeable and approached by widening eligibility criteria to preference trials and other non-randomized CCTs.However, even this approach resulted in a considerably low number of trials.Summed to the finding of only two trials reporting our primary outcomes, this review is further limited to the noncomparability of different questionnaires for patient satisfaction.Studies were also underpowered (modest sample sizes) for categorical outcomes.Major clinical heterogeneity also proscribes meta-analysis and thus contributes to the power-related issue.Our search strategy attempted to approach a wide series of potential sources for better sensitivity.Although we were initially limited to reports written in English, our search did not found non-English studies.Thus, language cannot be considered as a limitation of this review.This systematic review innovates by its patientcentered focus, which is uncommon in other reviews.However, it is notable that previous reviews found akin results for clinician/disease-centered outcomes.
Esposito, et al. 15 (2013) found similar survival and success rates for different loading methods.That review only considered clinical performance, and missed four of our six included reports given its last update timing. 15Finally, we extended the eligibility criteria to include preference trials, different from that review.Further three recent systematic reviews on IL's clinical outcomes 53-55 found a single CCT. 38Other reviews did not find CCTs comparing immediately loaded zygomatic implants to other loading protocols on similar fixtures either. 56,57ture trials are fundamental to compare IL to other loading approaches in the edentulous maxilla, and should consider zygomatic fixtures.Given that many patients can cope well with maxillary conventional dentures, and that the cost/complexity of IL may be quite high, recruitment in such trials can be quite slow.Multicenter RCTs can overcome those issues and timely reach a good sample size.Such tentative trial(s) should use standardized tools for outcome assessment at several recall visits -baseline up to at least 12 months, but focusing on short-term follow-up.Focus on patient-reported outcomes is imperative, given their fundamental role for clinical guidelines/recommendations. 17 Therefore, we present a systematic review of controlled clinical trials (CCT) comparing immediate versus early/delayed loading on implant-supported maxillary complete dentures, in terms of PROs and maintenance events/complications.This review was based on the following PICO question: in maxillary edentulous adults (P), is immediate loading (I) more effective than other loading protocols for full implant-supported prostheses (C) from the patient's perspective (O)?To reach a broad range of studies, we expanded this question to any treatment modality with complete implant support (i.e.fixed or removable).
Included studies should comply with the following criteria, grouped by design, participants, interventions, comparators and outcomes: Study design: experimental studies in humans comparing immediate loading to a control group (other loading protocols).The allocation of participants to one of the groups could be random (i.e.RCT) or not (nonrandomized CCT).Other designs (e.g.observational studies, one-arm trials) were not eligible.Participants: Adult patients with edentulous maxillae seeking implant-supported complete dentures.Interventions: Immediate-loaded, fully implantsupported complete dentures (IL): denture delivery until the 7 th day following implant insertion, 20 regardless of being the final or interim restoration.Dentures should be fixed or removable; in the latter MORRIS M, NADER SA, SOUZA RF case, they should receive complete support from implants.Eligible removable protocols include milled bars or telescopic attachment, given that the mucosa does not provide retention, stability or support.Comparators: Similar to the intervention, but with later delivery of a maxillary denture.Comparators were divided into (1) Early Loading (EL): loading between a week and two months after implant insertion; and (2) Conventional Loading (CL, also dubbed delayed loading): loading after more than two months after implant insertion.Outcome measures: Primary outcomes: general patient satisfaction with prostheses and oral healthrelated quality of life (OHRQoL), the most common PRO of studies on prosthodontics. 16Patient satisfaction could be graded by specific questions answered on categorical or quantitative scales; OHRQoL should be tested by validated questionnaires, including: Oral Health Impact Profile (OHIP), Oral Impacts on Daily Performance (OIDP), Geriatric Oral Health Assessment (GOHAI), and Dental Impact on Daily Living (DIDL), as well as their abbreviated versions.

( 2 )
Clinician-assessed implant-related parameters: implant success rate, marginal bone level, occurrence of mucositis and peri-implantitis, bleeding on probing (BOP), plaque index and probing depth.(3) Clinicianassessed performance of prostheses: success and survival rates, functional parameters like masticatory performance, technical complications like occlusal wear, screw loosening or fractured prosthetic components.
2019; results were limited to researches from 1999 onwards, due to the effective introduction of IL in the 1990s.Appendix 2 shows the search strategy used for MEDLINE via Ovid, which was adapted for each database.Given the search yield, we did not apply any filter or outcome-specific term.We also screened the list of references of included studies and reviews on immediate loading.The search was restricted to articles in English.Two authors (AA and RFS) scanned the titles and abstracts of all reports identified through the electronic searches independently.A 3 rd reviewer (SAN) was contacted as required to resolve disagreements.The same authors examined full-text versions of possible inclusions independently.

( 1 )
random sequence generation (selection bias), (2) allocation concealment (selection bias), (3) blinding (performance bias and detection bias), (4) incomplete outcome data (attrition bias), and (5) selective reporting (reporting bias), as well as (6) other sources.Each potential sources was classified as low, unclear or 2019;27:e20180600 4/15 Impact of immediately loaded implant-supported maxillary full-arch dental prostheses: a systematic review high.Moreover, the tool allows an overall classification of study risk of bias, i.e. any high-risk source renders the study as high risk of bias, whereas low-risk studies have all sources classified as such.Studies with any unclear source but no high-risk source were classified as moderate risk of bias.Summary measures and statistical analysis Most patient satisfaction and OHRQoL-related variables are continuous, and thus could be described according to their mean differences and 95% confidence interval (95%CI).Those included items answered on visual analogue scales (VAS) and summed results from Likert/ordinal scales.Similar strategies were used for other quantitative outcomes, including bone level changes.Dichotomous variables (e.g., frequency of prosthesis fracture, or occlusal wear: Yes/No) were described according to risk ratios (RR) with 95%CI.Whenever there were some issue regarding the unit of analysis for dichotomous variables (two or more event counts for the same participant), data was shown as cumulative incidence only.Inferences based on a 95%CI correspond to the adoption of a level of significance (α) of 0.05.The RevMan 5.3 software was used for plotting quality assessment and effect measures.

Figure 1 Figure 1 -
Figure 1 summarizes the search yield and study selection.We identified 1,052 reports by the electronic searches (duplicates excluded).Reading of titles and abstracts led to the exclusion of 98.1%, and to further appraisal of 20 full-text versions (1.9%).In turn, we included four trials reported by six manuscripts (two Spanish university clinic from 2008 to 2010, treated according to patient preferences.Both studies provided provisional acrylic maxillary fixed dentures immediately after implant insertion for IL.Provisional dentures were replaced by porcelain-fused-to metal (PFM) or metal-resin bridges after nearly 3 months.The other two included studies restricted their outcome assessment to clinical variables, and compared IL to CL.Both were conducted at university clinics and included further 64 participants (1 lost participant/arm).A non-randomized CCT in Italy compared IL on 4 to 6 implants to CL on 6 to 9 implants (loading time; IL: ≤24 h; CL: ~9 mo.). 39,40Recruitment happened between September 2005 and January 2006.Participants in the IL arm wore a transitional screw-retained acrylic fixed denture with a cast metal framework and without cantilevers during 4.5 months, followed by the definitive prostheses.Both arms received similar acrylic screw-retained definitive prostheses with one-tooth long cantilevers and cast metal frameworks.An RCT in Belgium also compared IL to CL (24 h versus 3 months) on a nonvariable number of six implants (surgery between February 2010 and December 2013). 41Both groups received detachable acrylic prostheses with cast metal frameworks, completely supported, stabilized and retained by SynCone telescopic abutments.Study ID Cannizzaro, et al. 38 (2008) Peñarrocha-Oltra, et al.
summarizes the quality assessment of the four included trials.Appendix 4 details the methodological quality assessment of individual trials.
38 (2008) performed a single assessment at the 12-month follow-up by asking whether patients were satisfied with overall treatment, indicated on a 5-point Likert scale.The RR of having participants more satisfied with IL after 12 months was 2.20 (95%CI: 1.01 to 4.79).
The use of reporting guidelines (e.g.SPIRIT and CONSORT) will lead to more transparent and comprehensive research methods, as well as trial registration in public databases (e.g.clinicaltrials.gov).ConclusionsThis review found modest evidence on the comparative performance of IL versus other loading regimens (CL and EL) for providing fully implantsupported maxillary dental prostheses.A limited number of trials suggest that patient satisfaction may be at least as good with IL, and show no major discrepancies regarding clinical complications.The selection of IL instead of CL or EL must rest onImpact of immediately loaded implant-supported maxillary full-arch dental prostheses: a systematic review solid practitioner's skills to provide such treatment and patient preferences.Evidence supports effective use of IL for fixed full prostheses on standard implants, given that no augmentation method is used.Patients seem at least as satisfied with IL, and clinical complications may be comparable.Comparative evidence on cases with unfavorable clinical features remains scant.

Table 1 -
Summary of the included study characteristicsImpact of immediately loaded implant-supported maxillary full-arch dental prostheses: a systematic review None of the four trials used grafting or other ridge augmentation procedures before implant insertion.

Table 2
summarizes the main findings of the four included trials, according to each outcome.

Table 2 -
in mm; positive values favor IL): Summary of outcome data from included studies (NR: not reported) * Distance between most coronal portion implant-bone contact area and coronal margin of implant collar; ** At the longest follow-up period/ total n comprises prosthetic complications + others J Appl Oral Sci.2019;27:e20180600 8/15