alveolar ridge preservation. a systematic review · 2014-10-17 · [15–17], xenografts, like...

Click here to load reader

Post on 14-Mar-2020

3 views

Category:

Documents

0 download

Embed Size (px)

TRANSCRIPT

  • REVIEW

    Alveolar ridge preservation. A systematic review

    Attila Horváth & Nikos Mardas & Luis André Mezzomo &Ian G. Needleman & Nikos Donos

    Received: 31 December 2011 /Accepted: 14 May 2012# Springer-Verlag 2012

    AbstractObjective The objective of this paper is to examine theeffect of alveolar ridge preservation (ARP) compared tounassisted socket healing.Methods Systematic review with electronic and hand searchwas performed. Randomised controlled trials (RCT), controlledclinical trials (CCT) and prospective cohort studies wereeligible.Results Eight RCTs and six CCTs were identified. Clinicalheterogeneity did not allow for meta-analysis. Average changein clinical alveolar ridge (AR) width varied between −1.0and −3.5±2.7 mm in ARP groups and between −2.5and −4.6±0.3 mm in the controls, resulting in statisticallysignificantly smaller reduction in the ARP groups in five outof seven studies. Mean change in clinical AR height variedbetween +1.3±2.0 and −0.7±1.4 mm in the ARP groups andbetween −0.8±1.6 and −3.6±1.5 mm in the controls. Heightreduction in the ARP groups was statistically significantly less

    in six out of eight studies. Histological analysis indicatedvarious degrees of new bone formation in both groups. Somegraft interfered with the healing. Two out of eight studiesreported statistically significantly more trabecular bone for-mation in the ARP group. No superiority of one technique forARP could be identified; however, in certain cases guidedbone regeneration was most effective. Statistically, signifi-cantly less augmentation at implant placement was neededin the ARP group in three out of four studies. The strength ofevidence was moderate to low.Conclusions Post-extraction resorption of the AR might belimited, but cannot be eliminated by ARP, which at histolog-ical level does not always promote new bone formation. RCTswith unassisted socket healing and implant placement in theARP studies are needed to support clinical decision making.Clinical relevance This systematic review reports not only onthe clinical and radiographic outcomes, but also evaluates thehistological appearance of the socket, along with site specificfactors, patient-reported outcomes, feasibility of implantplacement and strength of evidence, which will facilitate thedecision making process in the clinical practice.

    Keywords Tooth extraction . Bone resorption . Implant sitedevelopment . Bone substitute . Bone regeneration .

    Human histology

    Introduction

    Periodontal disease, periapical pathology and mechanical trau-ma often result in bone loss prior to tooth removal [1]. Further-more, traumatic extraction has also been associated withadditional loss of bone. In the healing phase after extraction,alveolar bone undergoes additional atrophy as a result of thenatural remodelling process [2–7]. This begins immediately

    A. Horváth :N. Mardas : L. A. Mezzomo : I. G. Needleman :N. Donos (*)Unit of Periodontology, Department of Clinical Research,UCL Eastman Dental Institute,256 Gray’s Inn Road,London WC1X 8LD, UKe-mail: [email protected]

    A. HorváthDepartment of Periodontology, Semmelweis University,Budapest, Hungary

    L. A. MezzomoPontifical Catholic University of Rio Grande do Sul,Porto Alegre, Brazil

    I. G. NeedlemanInternational Centre for Evidence-Based Oral Health,UCL Eastman Dental Institute,London, UK

    Clin Oral InvestDOI 10.1007/s00784-012-0758-5

  • after extraction and may result in up to 50 % resorption of thealveolar ridge (AR) width even in 3 months [1]. Post-extractionAR resorption may have an impact on dental implant place-ment, since sufficient vertical and horizontal volume of alveolarbone should ideally be present at the site of insertion [8].

    Alveolar ridge preservation (ARP) procedures have beenintroduced to maintain an acceptable ridge contour in areasof aesthetic concern, as well as to prevent alveolar ridgeatrophy and maintain adequate dimensions of bone in orderto facilitate implant placement in prosthetically driven posi-tions [9, 10]. Several methods have already been investigat-ed for ARP in preclinical models [11–14] and clinicalstudies, such as socket grafting with autogenous bone[15], demineralised freeze-dried bone allograft (DFDBA)[15–17], xenografts, like deproteinized bovine-bonemineral (DBBM) [18], alloplasts [19] and bone morpho-genic proteins (BMP) [20]. Guided bone regeneration(GBR) with or without bone grafts has also been evaluated[9, 10, 21–25].

    Although some of the above bone substitutes wereable to limit the resorption of post-extraction alveolarridge up to a certain extent, the quality of the newtissue in the socket varied broadly. The remnants ofthe grafts often interfered with the normal healing pro-cess in line with preclinical results [15–17, 26]. Anumber of review articles on ARP have been publishedin the last decade [27–32]. However, a systematic as-sessment of the nature and quality of the newly formedtissue alongside methodological quality and risk of biasof the studies has not been carried out. Furthermore,non-controlled prospective and retrospective studies aswell as case series were also included in most of theprevious reviews without the comparison to the controlgroup of unassisted socket healing [33–36].

    Therefore, the objective of the present systematic reviewwas to investigate the effect of ridge preservation on theresidual alveolar ridge dimensions and on histological char-acteristics, compared to unassisted socket healing.

    Methods

    Prior to commencement of the study, a detailed protocol wasdeveloped and agreed upon by the authors based on theCochrane Collaboration guidelines and previous reviewspublished by our group [37–41].

    Focused question

    Following tooth/root extraction in humans, what is the effectof ridge preservation on the residual alveolar ridge dimensionand on histological characteristics, compared to unassistedsocket healing?

    Definition

    Whilst ‘socket preservation’ has widely been employed todepict a certain procedure, we believe that the objective ofthese interventions is to preserve the dimension of the AR.Therefore, we have used the term ‘Alveolar Ridge Preser-vation’ to define such procedures.

    Types of studies

    Longitudinal prospective studies were included, i.e. RCTs,CCTs and cohort studies with control group.

    Populations of studies

    Healthy individuals, without any age limit, who underwentany type of ridge preservation following permanent toothextraction, were included. Smokers and patients with historyof periodontal disease were not excluded. The minimumnumber of subjects per group was five. However, no limitwas set for study follow-up period.

    Types of interventions

    Test groups

    Studies reporting on any of the following types of interven-tions were included: socket grafting (autograft, allograft,xenograft, alloplastic materials); socket sealing (soft tissuegrafts); GBR (resorbable/non-resorbable barriers); biologicalactive materials (growth factors) and combinations of theabove techniques/materials.

    Control groups

    The control groups of the included studies comprised emptysockets, i.e. unassisted socket healing.

    Outcome variables

    The primary outcomewas the change in oro-facial (horizontal)and apico-coronal (vertical) AR dimensions. Secondary out-comeswere the following: (1) change in buccal plate thickness;(2) bone volume alteration following extraction; (3) complica-tions; (4) histological healing characteristics; (5) site eligibilityfor placement of an adequate size dental implant with orwithout further augmentation; (6) patient-reported outcomes,such as quality of life and (7) health economics.

    Risk of bias and methodological quality assessment

    In order to evaluate the methodological quality and risk ofbias of individual studies, we used a combination of

    Clin Oral Invest

  • parameters from the Cochrane Collaboration and Consoli-dated Standards of Reporting Trials (CONSORT) statement.The following parameters were assessed and taken intoconsideration in the final analysis: sample size calculation,statement of eligibility criteria, ethics approval, informedconsent, baseline homogeneity, randomisation method, allo-cation concealment, masking, calibration, follow up, protocolviolation, method of statistics, unit of analysis, CONSORTimplementation, International Standard Randomised Con-trolled Trial Number Register (ISRCTN) and funding disclo-sure. Methodology unique to RCTs was not assessed in CCTs,i.e. randomisation and concealment of allocation.

    Randomisation was accepted as adequate, in case theallocation sequence was correctly generated either bycomputer, toss of a coin, throwing dice, etc. Quasirandomisation, e.g. birth dates, hospital numbers werenot accepted. Adequacy of allocation concealment wasaccepted if the sequence was concealed, until interven-tion was assigned (e.g. in sequentially numbered andsealed opaque envelopes, remote computer or centraltelephone). Statistical analysis was judged as adequateif appropriate statistical method was selected to accom-modate to the characteristic of the each individual data(e.g. number of groups and investigated categories, sizeof samples, normally distributed or skewed data, para-metric or non-parametric, paired or unpaired, numericalor categorical variables). Statistical significance was ac-cepted in case of confidence interval (CI) >95 % (p<0.05), while ‘statistically highly significant’ referred toCI>99.9 % (p

  • interface. A three-stage selection of the resulted hits wasperformed independently and in duplicate by two reviewers(AH and LAM). In order to reduce errors and bias, a cali-bration exercise was performed with the first 500 titles,resulting in 96.4 % agreement. In case of disagreement atthe title selection stage, the trial was included in the abstractstage. At the abstract and full text selection any disagree-ments between the above reviewers were resolved by dis-cussion. If unresolved, a third reviewer (NM) was involvedfor arbitration. The reasons for exclusion were recordedeither in the Reference Manager (abstract stage) or in aspecific data extraction form (full text stage). The level ofagreement was determined by Kappa score calculation.

    Research synthesis

    Studies were grouped by research design and their chiefcharacteristics. Outcomes were recorded in evidence tables.In view of the marked heterogeneity, no meta-analysis wasconducted. Instead, a narrative synthesis was undertaken.

    Results

    Search sequence

    The electronic search yielded 6,216 relevant hits after re-moval of duplicates (Fig. 1). Subsequently, 157 titles wereselected for the abstract stage. Following investigation ofthe abstracts, 42 articles qualified for full text evaluation.Four extra papers were then added as a result of the handsearch. Assessment of these articles resulted in the following

    14 publications eligible for the review [17, 19–21, 23–25,42–48]. The excluded full text papers along with the reasonsfor exclusion are listed in Table 1. The most typical reasonsfor exclusion were lack of control group with unassistedsocket healing; use of retrospective design; assessment ofdimensional changes of the AR only on periapical two-dimensional radiographs, or on casts taken from soft tissuelevel; and surgical removal of third molars.

    The Kappa score for agreement between the reviewers (AH,LAM) at the abstract and full text selection level, was 0.96 and0.90, respectively, indicating a high level of agreement.

    Study characteristics

    In the 14 included articles (eight RCTs and six CCTs) theefficacy of ARP techniques was evaluated clinically bymeans of direct measurements of the residual alveolar ridgedimensions during re-entry procedures, radiographically bymeans of computer tomography or histologically from tre-phine biopsies taken at re-entry during osteotomies forimplant placement (Tables 3 and 4). No cohort studies wereindentified. Limited data were reported on confoundingfactors, such as periodontitis, smoking, systemic diseaseand medication. The extraction site distribution was fairlyheterogeneous. In some studies ARP was performed only inmaxillary anterior sockets [42, 46, 47], whereas such restric-tion was not employed in other studies. The residual bonevolume around the investigated sockets, e.g. the presence/absence and width of the buccal bone plate varied fromseverely compromised [20, 46], to completely intact, buccalbone (Table 3) [17, 21, 42].

    Intervention characteristics

    With regard to the techniques or materials used for ARP, theincluded studies were grouped into three categories (Table 3);

    1. Bone grafts/substitutes2. GBR3. Biological active materials.

    In the majority of the included studies, various bonegrafts were utilised, such as autologous bone marrow [47],plasma rich in growth factor (PRGF) with or without autol-ogous bone [43], DFDBA [17], DBBM [46], calcium sul-phate hemihydrates [42, 45] and bioactive glass [17].Alloplastic polyglycolide/polylactide (PGPL) sponge wasalso employed [19, 48]. GBR technique was applied usingnon-resorbable expanded polytetrafluoroethylene (e-PTFE)[24] or resorbable (PGPL) [25] barrier. Resorbable collagenmembrane was also employed in combination with FDBA[23] or corticocancellous porcine bone [21]. Biological ac-tive material, namely bone morphogenic protein (rhBMP-2)was used on a collagen sponge carrier in one study [20].

    Electronic search6.216 titles

    Included publications14

    Relevantabstracts

    Full-text analysis 45

    Relevant full-texts42

    6,059 Excluded based

    on the title

    115Excluded based on the abstract

    32Excluded based on the full-text

    3 Included as a result of hand search

    Kappa score 0.96

    Kappa score 0.90

    1 Included as a result of final search

    Fig. 1 Flow chart of the screening process

    Clin Oral Invest

  • Table 1 List of excluded full text papers and reasons for exclusion

    First author(year of publication)

    Journal Reasons for exclusion

    Bianchi (2004) Int J Periodont Rest Dent Retrospective analysis

    Single-arm of the included Fiorellini et al. (2005)

    Bolouri (2001) Comp Cont Educ Dent Reported on optical density on two-dimensional radiographs

    Brawn (2007) Impl Dent Case report

    Brkovic (2008) J Can Dent Assoc Case report

    Carmagnola (2003) Clin Oral Impl Res Lack of real control group, resembles to a retrospective analysis(extreme difference in follow-up period between tests and controls.T1: 4 months; T2: 7 months; C: 1-15 years, mean: 7.8 years)

    Cranin (1988) J Biomed Mat Res Case series without control group

    De Coster (2009) Clin Impl Dent Relat Res Case series

    Retrospective study as stated by the authors in the discussion

    Healing period varied between 1.5 months and 1.5 years

    Neither histomorphometry nor clinical or radiographic measurementsreported in the results

    Graziani (2008) J Cranofac Surg Extraction of fully impacted third molars

    Linear measurements on OPG

    Gulaldi (1998) Oral Surg Oral MedOral Pat Oral Rad End

    Extraction of fully impacted third molars

    Linear measurements on OPG and scintigraphy

    Primary outcome was to analyze bone metabolism

    Heberer (2008) Clin Oral Impl Res Case series without control group

    Hoad-Reddick (1994) Eur J Prosth Rest Dent Two-dimensional linear measurements obtained from OPG and cephalometry

    Lack of defined landmarks

    Surgical procedure was not described

    Hoad-Reddick (1999) Eur J Prosth Rest Dent Description of a method for measurements on casts

    Neither socket preservation procedure nor the results were described.Soft tissue punch technique only

    Howell (1997) Int J Periodont Rest Dent Case series without control group

    Jung (2004) Int J Periodont Rest Dent Case series without control group

    Primary outcome was soft tissue healing

    Kangvonkit (1986) Int J Oral Maxillofac Surg Based on OPG and lateral cephalogram only

    Evaluation method remains unclear

    Primary outcome was the biocompatibility of HA cones

    Karapataki (2000) J Clin Periodontol Extraction of fully impacted third molars

    Primary outcome was to assess the periodontal status ofsecond molars after extraction of third molars

    Kerr (2008) J Periodontol No biomaterials were used to preserve the ridge dimensions,therefore did not address the focused question

    Kwon (1986) J Oral Maxillofac Surg Based on OPG and lateral cephalogram only

    Evaluation method remains unclear

    Lack of description of the measurement methods

    Molly (2008) J Periodontol Control group was covered by an e-PTFE membrane,thus lack of unassisted control sockets

    Munhoz (2006) Dento Maxillofac Radiol Extraction of fully impacted third molars

    Two-dimensional evaluation of periapical radiographs

    Norton (2002) Int J Oral Maxillofac Impl Case series without control group

    Resembles to a retrospective design(healing period ranged from 3 to 11 months)

    Page (1987) J Oral Maxillofac Surg Case report

    Pape (1988) Deutsche ZahnarztlicheZeitschrift

    Augmentation of a resorbed ridge

    Clin Oral Invest

  • None of the included studies used the socket sealing tech-nique. Primary flap closure was achieved in 9 out of 14studies, while the sockets left uncovered in the rests. Varioustypes and amounts of antibiotics and antiseptic rinses wereadministered for different duration in studies reporting onpostoperative care. Finally, average healing period rangedfrom one to nine months.

    Outcome characteristics

    Clinical outcomes

    Eight out of the 14 included studies investigated the efficacyof various ARP techniques to preserve the pre-extractionridge dimensions using intra-surgical hard tissue measure-ments taken during re-entry procedure [19, 21, 23–25, 42,44, 47]. In these studies, ARP was performed in 137 socketsof 119 patients and compared to 120 sockets that left to healwithout any treatment in a total of 92 patients (Table 3).

    Bone ‘graft’ Four studies evaluated changes in AR dimen-sions following grafting of the socket. Two studies were RCTs[42, 47] and two were CCTs [19, 44]. Healing time variedfrom 3 to 6 months [19, 42, 44, 47].

    The horizontal (bucco-lingual) changes of the alveo-lar ridge were assessed in three studies [42, 44, 47].The AR reduced in width from baseline to re-entrybetween −1.0 mm and −3.5±2.7 mm following ARP(p

  • 25] and one was CCT [24]. Healing time varied between 4and 9 months.

    Horizontal (bucco-lingual) changes of the AR wereassessed in all four studies. AR width reduction from base-line to re-entry varied between −1.2±0.9 mm and −2.5±1.2 mm in the GBR-treated sockets and between −2.6±2.3 mm and −4.6±0.3 mm in the control groups. With theexception of one study [23], a statistically significantlysmaller reduction of the alveolar ridge width was observedwhen e-PTFE [24], PGPL [25], or collagen membranes incombination with xenograft [21] were used.

    All the four studies investigated the mean change in ARheight at the mid-buccal aspect. The AR height changedfrom baseline to re-entry between +1.3±2.0 mm and −0.7±1.4 mm in the ARP groups and between −0.9±1.6 mmand −3.6±1.5 mm in the control groups. The resorption inthe ARP group was not statistically significant in three outof four studies [23–25]. All studies reported a statisticallysignificantly less post-extraction reduction in AR heightwhen the socket was treated by GBR compared to unassistedhealing.

    Vertical dimension changes at mesial and distal aspectsof the socket were measured in two studies [21, 23]. Theobserved differences between baseline and re-entry were notstatistically significant in both groups. In one out of the twostudies the height reduction was statistically significantlysmaller in the test group compared to the control [23].

    Two studies captured data on the socket fill [24, 25] andreported statistically significant socket fill in both groupsbetween baseline and re-entry, as well as between tests andcontrols.

    No data were found on either initial buccal plate thick-ness or alteration of bone volume. However, one studymeasured the buccal bone thickness loss and reported sta-tistically significantly less reduction in the ARP group [47].

    Radiographic measurements

    Two RCTs, reporting on 3D radiographic assessment, metthe inclusion criteria [20, 46]. The healing time varied from1 to 4 months. In one study, where the post-extraction socketwas grafted with a radiopaque material (DBBM), treatmentresulted in significantly less reduction in radiographic ARheight compared to unassisted socket healing [46]. The testgroup in the other study, where the higher concentration(1.5 mg/ml) of RhBMP-2 was utilised [20], resulted in amean increase of the radiographic AR width by 3.27±2.53 mm at the most coronal part, compared to the 0.57±2.56 mm increase in the group of unassisted healing. ARheight was reduced by 0.02±1.2 mm in the same test groupand by 1.17±1.23 mm in the control group (Table 3). Thedifferences between test and control were statisticallysignificant.

    Histological results

    Eleven studies carried out a histological analysis based ontrephine biopsies retrieved at re-entry [17, 19–21, 23, 42,43, 45–48]. Seven studies were RCTs [17, 20, 21, 23, 46,47] and four were CCTs [19, 43, 45, 48]. In these studies,ARP was performed in 181 sockets of 158 patients andcompared to 149 sockets that left to heal without anytreatment in 131 patients (Table 4). Only two out ofeight studies reported statistically significantly higher tra-becular bone volume following ARP in comparison to unas-sisted socket healing [21, 42] and two studies reportedstatistically significantly more connective tissue in the post-extraction socket when no ARP was performed [17, 21]. Onthe contrary, one study reported more vital bone in theunassisted socket healing group compared to the ARPgroup [23]. None of the differences of the investigatedhistomorphometric parameters reached statistical signifi-cance in other studies.

    Bone ‘grafts’ Eight studies evaluated histologically thehealing of post-extraction sockets following the applicationof some type of bone grafts/substitutes [17, 19, 42, 43,45–48]. Four studies were RCTs [17, 42, 46, 47] and fourwere CCTs [19, 43, 45, 48]. New mineralised bone wasobserved at various levels in all studies in both ARP andcontrol groups in a healing period from 2.5 to 8 months.Connective tissue occupied a portion of the socket in bothgroups. When DFDBA, bioactive glass or DBBM wereused, the graft particles were embedded either in new boneor in connective tissue. In most studies, there was no sig-nificant difference in the type of healing, or amount of boneformation between bone grafts and unassisted sockethealing.

    GBR in combination with graft GBR in combination withgraft was utilised in two RCTs. ARP with a collagen mem-brane and deproteinized porcine bone resulted in statisticallysignificantly higher new bone and lower connective tissueformation after 7 to 9 months of healing in comparison tounassisted socket healing [21]. However, residual graftmaterials were present in the ARP biopsies. FDBA andcollagen membrane resulted in similar amounts of new boneformation to untreated sockets, although more vital bonewas observed in the untreated sockets at 4 to 6 months ofhealing (p>0.05) [23].

    Biological active material RhBMP-2 in a collagen spongecarrier was completely resorbed at 4 months following ARPregardless of the concentration of the growth factor [20].Mineralised tissue was found and trabecular bone formationwas noticed in two third of both the test and control biopsiesin the RCT.

    Clin Oral Invest

  • Adverse events, complications

    Adverse events were reported in six RCTs [17, 20, 21, 25,42, 47] and four CCTs [19, 24, 44, 48] including oedema,pain, erythema and membrane exposure/infection. In twostudies, more adverse events, i.e. oedema, erythema [20] ormembrane exposure [24] were observed in the ARP groupcompared to the natural socket healing. No comparisonbetween tests and controls were reported in the other studies(Table 3).

    Feasibility of implant placement

    Seven studies [17, 19, 23, 42, 45, 46, 48] reported thatimplant placement in the previous sockets were successful,but no differences between the ARP and untreated sites wererevealed. The outcome of implant placement remained un-clear in one article [43] and only re-entry without implanta-tion was performed in three trials [24, 44]. Four studiesreported the need of further augmentation at the stage ofimplant placement. Three of them favoured the ARP groupover the controls, since less [20] or no sites [21, 47] in theARP group presented with residual dehiscence or fenestra-tion defects around the inserted implants (Table 3).

    Patient-reported outcome and health economics

    No data were found for patient-reported outcome measuresor health economic evaluation.

    Quality assessment

    Considerable heterogeneity was found among the studies interms of methodological quality. Detailed description of thequality assessment of the included studies is presented inTable 2. Among the 14 included controlled studies, eightwere randomised [17, 20, 21, 23, 25, 42, 46, 47] although infour of them the randomisation technique was not reported[20, 42, 46, 47]. None of the RCTs reported the method ofallocation concealment. Masking of the examiner wasreported at the clinical level in two out of eight [23, 25], atradiological level in one out of two [20] and at histologicallevel in four out of 11 studies [17, 21, 42, 43]. Examinercalibration was declared in three papers [20, 23, 42], whilstinclusion and exclusion criteria were defined in seven pub-lications [17, 21, 23, 42, 43, 46, 47]. Apart from threestudies [21, 43, 46] all the other reported the approval ofthe ethical committee. Three studies were funded by indus-try [17, 20, 44], two studies by academic institution [45, 48]and the remaining nine did not report the source of funding.

    Nine trials implemented patient-based analysis [20, 21,23–25, 42, 44, 47, 48], whilst the extraction site served asunit of analysis in the rest of the five investigations [17, 19,

    43, 45, 46]. Sample size calculations were reported only inthree studies [20, 23, 42], although with insufficient data toevaluate the validity of the calculations. Statistical analysiswas appropriately carried out and described in one studyonly [47]. Appropriate statistics were either not carried out[17, 19–21, 43, 45, 46], or the reported data were insuffi-cient to determine the validity [23–25, 42, 43, 48]. In addi-tion, no RCTs were either registered with ISRCTN orreported using the CONSORT guidelines (Table 3).

    Risk of bias

    Four studies were classified as moderate risk of bias [17, 21,23, 25] and the rest were categorised as high risk of bias(Table 2).

    Discussion

    Key findings

    This systematic review has demonstrated that different ARPtechniques do not totally eliminate post-extraction alveolarridge resorption or predictably promote new bone forma-tion. However, the reduction in ridge width and heightfollowing ARP may be less than that which occurs follow-ing natural socket healing. The clinical data suggest that thehorizontal ridge contraction was most successfully limitedin the two studies applying GBR without additional bonegrafts [24, 25], whereas the vertical shrinkage was mostefficiently limited by employing GBR with additional bonegraft [21, 23].

    Strengths of the review

    The present systematic review was limited to randomisedcontrolled trials, controlled clinical trials and prospectivecohort studies with a control group of empty untreatedsockets. Furthermore, the inclusion criteria of our systematicreview were based on the fact that the clinical merit ofapplying the different ARP techniques could only be vali-dated, if the clinical and histological outcomes following theapplication of a technique are superior to that of unassistedsocket healing.

    In comparison to the previous systematic reviews [28,32] the present review has evaluated the histological char-acteristics of the alveolar socket healing with or withoutARP. The amount and the quality of the newly formedosseous tissues in the socket area are essential, especiallywhen the justification of ARP is to facilitate the placementof a dental implant in the position of a previously extractedtooth. It is doubtful, whether an ARP technique should beclaimed successful, if it only preserves the external contour

    Clin Oral Invest

  • Tab

    le2

    Qualityassessmentof

    theinclud

    edstud

    ies

    Study

    QualityCriteria

    Estim

    ated

    risk

    ofbias

    Firstauthor

    Randomisation

    Masking

    Calibratio

    nElig

    ibility

    Criteria

    Follow

    upEthical

    considerations

    Funding

    Statistical

    analysis

    Miscellaneous

    Yearof

    publication

    1.Randomised

    1.Therapist

    1.Intra-exam

    iner

    1.Inclusion

    criteriadefined

    1.Percentage

    ofcompleted

    follo

    wups

    1.Ethicsapproval

    Sourceof

    Funding

    1.Appropriate

    samplesize

    calculationandpower

    1.Com

    parable

    experimentalgroups

    2.CONSORT

    implem

    ented

    2.Inform

    edconsent

    3.ISRCTN

    registered

    2.Unitof

    analysis

    4.Other

    comments

    3.Appropriate

    statisticsapplied

    2.Exclusion

    criteriadefined

    2.Adequate

    correctio

    n

    2.Patient

    2.Inter-exam

    iner

    2.Adequate

    sequence

    generatio

    n3.

    Examiner

    Type

    4.Statistician

    3.Allo

    catio

    nconcealm

    ent

    Reference

    number

    4.Concealment

    adequate

    Aim

    etti

    1.Yes

    1.N/R

    1.Yes

    (histo),

    N/R

    (clin

    )1.

    Yes

    1.N/R

    1.Yes

    N/R

    1.Insufficient

    data

    todeterm

    ine

    1.Yes

    High

    2.N/R

    2009

    2.N/R

    2.N/R

    2.N/A

    2.Yes

    2.N/A

    2.Yes

    2.Patient

    3.N/R

    3.Insufficient

    data

    todeterm

    ine

    RCT

    3.N/R

    3.Yes

    (histo),

    N/R

    (clin

    ).#42

    4.N/A

    4.N/R

    *Reportedas

    ‘doubleblind’

    Anitua

    1.Yes

    (btw

    T-C)

    No(w

    ithin

    T)

    1.N/R

    1.N/R

    1.Yes

    1.100%

    1.N/R

    N/R

    1.N/R

    1.N/R

    High

    1999

    2.N/A

    2.N/R

    2.N/A

    2.Yes

    2.Yes

    2.Yes

    2.Patient

    +site

    2.N/R

    CCT

    3.N/R

    3.Yes

    3.Nostatistical

    analysis

    was

    carriedout

    3.N/R

    #43

    4.N/A

    4.N/R

    4.Atsevere

    defects

    autogenous

    bone

    was

    addedto

    PRGF.

    Different

    healingperiods.

    Barone

    1.Yes

    1.N/R

    1.N/R

    1.Yes

    1.100%

    1.N/R

    N/R,declared

    noconflict

    ofinterest

    1.N/R

    1.Yes

    Moderate

    2008

    2.Yes

    2.N/R

    2.N/A

    2.Yes

    2.Yes

    2.Yes

    2.Patient

    2.N/R

    RCT

    3.N/R

    3.Yes

    (histo),

    N/R

    (clin

    )3.

    No

    3.N/R

    #21

    4.N/A

    4.N/R

    4.Different

    healing

    periods.

    Cam

    argo

    N/A

    1.N/R

    1.N/R

    1.Yes

    1.100%

    1.Yes

    Industry

    1.N/R

    1.N/R

    High

    2000

    2.N/R

    2.N/A

    2.Yes

    2.Yes

    2.Yes

    2.Patient

    2.N/R

    CCT

    3.N/R

    3.Insufficient

    data

    todeterm

    ine

    3.N/R

    #44

    4.N/R

    Fiorellini

    1.Yes

    1.N/R

    1.N/R

    1.No

    1.100%

    1.Yes

    Industry

    1.Insufficient

    data

    1.N/R

    High

    2005

    2.N/R

    2.N/R

    2.Yes

    2.No

    2.Unclear

    2.Yes

    todeterm

    ine

    2.N/R

    RCT

    3.N/R

    3.Yes

    (CTscans)

    2.Patient

    3.N/R

    #20

    4.N/A

    4.N/R

    *Reported

    as‘double

    blind ’

    3.No

    4.Standardisatio

    nof

    CTscansN/R.

    Final

    number

    ofsockets,patients

    remainunclear.

    Froum

    1.Yes

    1.N/R

    1.N/R

    1.Yes

    1.100%

    1.Yes

    Industry

    1.N/R

    1.N/R

    Moderate

    2002

    2.Yes

    2.N/R

    2.N/A

    2.Yes

    2.Unclear

    2.Yes

    2.Site

    2.N/R

    RCT

    3.N/R

    3.Yes

    3.No

    3.N/R

    #17

    4.N/A

    4.N/R

    4.Different

    healingperiods.

    Enrolmentof

    sitesof

    subjectsinconsistent.

    Clin Oral Invest

  • Tab

    le2

    (con

    tinued)

    Study

    QualityCriteria

    Estim

    ated

    risk

    ofbias

    Firstauthor

    Randomisation

    Masking

    Calibratio

    nElig

    ibility

    Criteria

    Follow

    upEthical

    considerations

    Funding

    Statistical

    analysis

    Miscellaneous

    Yearof

    publication

    1.Randomised

    1.Therapist

    1.Intra-exam

    iner

    1.Inclusion

    criteriadefined

    1.Percentage

    ofcompleted

    follo

    wups

    1.Ethicsapproval

    Sourceof

    Funding

    1.Appropriate

    samplesize

    calculationandpower

    1.Com

    parable

    experimentalgroups

    2.CONSORT

    implem

    ented

    2.Inform

    edconsent

    3.ISRCTN

    registered

    2.Unitof

    analysis

    4.Other

    comments

    3.Appropriate

    statisticsapplied

    2.Exclusion

    criteriadefined

    2.Adequate

    correctio

    n

    2.Patient

    2.Inter-exam

    iner

    2.Adequate

    sequence

    generatio

    n3.

    Examiner

    Type

    4.Statistician

    3.Allo

    catio

    nconcealm

    ent

    Reference

    number

    4.Concealment

    adequate

    Guarnieri

    N/A

    1.N/R

    1.N/R

    1.Yes

    1.N/R

    1.Yes

    Governm

    ent;

    institu

    tion

    1.N/R

    1.N/R

    High

    2004

    2.N/R

    2.N/A

    2.No

    2.N/A

    2.Yes

    2.Site

    2.N/R

    CCT

    3.N/R

    3.No

    3.N/R

    #45

    4.N/R

    Iasella

    1.Yes

    1.N/R

    1.Yes

    1.Yes

    1.100%

    1.Yes

    N/R

    1.Insufficient

    data

    todeterm

    ine

    1.Yes

    Moderate

    2003

    2.Yes

    2.N/R

    2.N/A

    2.Yes

    2.Yes

    2.Yes

    2.Patient

    2.N/R

    RCT

    3.N/R

    3.Yes

    3.Insufficient

    data

    todeterm

    ine

    3.N/R

    #23

    4.N/A

    4.N/R

    Lekovic

    N/A

    1.N/R

    1.N/R

    1.No

    1.70%

    (premature

    exposure

    ofe-PTFE

    barrierin

    3/10)

    1.Yes

    N/R

    1.N/R

    1.Yes

    High

    1997

    2.N/R

    2.N/A

    2.No

    2.N/R

    2.Patient

    2.N/R

    CCT

    3.N/R

    2.Yes

    3.Insufficient

    data

    todeterm

    ine

    3.N/R

    #24

    4.N/R

    Lekovic

    1.Yes

    1.N/R

    1.N/R

    1.No

    1.100%

    1.Yes

    N/R

    1.N/R

    1.Yes

    Moderate

    1998

    2.Yes

    2.N/R

    2.N/A

    2.No

    2.Yes

    2.Yes

    2.Patient

    2.N/R

    RCT

    3.N/R

    3.Yes

    3.Insufficient

    data

    todeterm

    ine

    3.N/R

    #25

    4.N/A

    4.Yes

    Nevins

    1.Yes

    1.N/R

    1.N/R

    1.Yes

    1.100%

    1.N/R

    N/R

    1.N/R

    1.Yes

    High

    2006

    2.N/R

    2.N/R

    2.N/A

    2.Yes

    2.Yes

    2.N/R

    2.Site

    2.N/R

    RCT

    3.N/R

    3.N/R

    3.No

    3.N/R

    #46

    4.N/A

    4.N/R

    4.Standardisationof

    CT

    scansN/R.T

    estm

    aterial

    radiopaque.D

    ifferent

    healingperiods.

    Pelegrine

    1.Yes

    1.N/R

    1.N/R

    1.Yes

    1.100%

    1.Yes

    N/R

    1.N/R

    1.N/R

    High

    2010

    2.N/R

    2.N/R

    2.N/A

    2.Yes

    2.Yes

    2.Yes

    2.Patient

    2.N/R

    RCT

    3.N/R

    3.N/R

    3.Yes

    3.N/R

    #47

    4.N/A

    4.N/R

    Serino

    N/A

    1.N/R

    1.N/R

    1.Yes

    1.80%

    1.Yes

    N/R

    1.N/R

    1.N/R

    High

    2003

    2.N/R

    2.N/A

    2.No

    2.Unclear

    2.Yes

    2.Site

    2.N/R

    CCT

    3.N/R

    3.No

    3.N/R

    #19

    4.N/R

    4.Molarsonly

    inT.

    Clin Oral Invest

  • of the AR, but the newly formed tissue is of inferior qualityand quantity (percentage of matured trabecular bone) towhat is normally achieved following a tooth extraction.

    Finally, the quality of the included studies has also beenmeticulously assessed in this review. Such a quality evalu-ation of the retrieved data is essential to estimate the sourceand magnitude of potential bias that may lead to delusiveconclusions.

    Strength of evidence—risk of bias

    The quality assessment of the included studies in this sys-tematic review revealed that none of the trials have qualifiedfor a low risk of bias category. Ten out of the 14 studiespresented with high risk of bias thus their results must beevaluated with caution. The lack of clear reporting of re-search methodology elements, such as adequate randomiza-tion and concealment and/or masking of the therapist andthe examiner were among the primary reasons for the highrisk of bias [49]. We did not contact authors for clarificationof unclear methodology. Therefore, it is possible that actualstudy conduct was better than that reported in the publica-tion. Statistical considerations played important role as well,since appropriate analytical statistics was completed andreported merely in one study [47]. Power calculation wasconducted in three trials only [21, 23, 42], nevertheless thereported data were insufficient to determine the validity ofthe calculation.

    Dimensional changes and histological characteristics

    Sufficient ridge width and height have been considered asone of the key requirements for successful implant therapyand for the establishment of an aesthetically pleasing emer-gence profile at fixed partial dentures [8, 50, 51]. Therefore,the alterations in oro-facial (horizontal) and apico-coronal(vertical) AR dimensions were selected as the primary out-comes of the present review. Direct intra-surgical measure-ments on the AR at re-entry are considered as the mostprecise method to evaluate the bone volume changes fol-lowing ARP. It is desirable though to establish and validate asurrogate measure that avoids the need for re-entry surgery,while providing the clinician with a reliable measure. Two-dimensional radiographs, such as periapical or panoramicradiographs, are not ideal to estimate the 3D changes of theAR [52]. Also, measurements of the alveolar mucosa levelor study casts incorporate not only the alveolar bone, butalso the overlaying soft tissue. For these reasons onlystudies performing clinical or 3D radiographic evaluationof hard tissue were included in this review. Cone-beamcomputerised tomography (CBCT) appears to offer a validtechnique to assess alveolar ridge changes, with newer mod-els greatly reducing radiation exposure [53]. However, aTa

    ble

    2(con

    tinued)

    Study

    QualityCriteria

    Estim

    ated

    risk

    ofbias

    Firstauthor

    Randomisation

    Masking

    Calibratio

    nElig

    ibility

    Criteria

    Follow

    upEthical

    considerations

    Funding

    Statistical

    analysis

    Miscellaneous

    Yearof

    publication

    1.Randomised

    1.Therapist

    1.Intra-exam

    iner

    1.Inclusion

    criteriadefined

    1.Percentage

    ofcompleted

    follo

    wups

    1.Ethicsapproval

    Sourceof

    Funding

    1.Appropriate

    samplesize

    calculationandpower

    1.Com

    parable

    experimentalgroups

    2.CONSORT

    implem

    ented

    2.Inform

    edconsent

    3.ISRCTN

    registered

    2.Unitof

    analysis

    4.Other

    comments

    3.Appropriate

    statisticsapplied

    2.Exclusion

    criteriadefined

    2.Adequate

    correctio

    n

    2.Patient

    2.Inter-exam

    iner

    2.Adequate

    sequence

    generatio

    n3.

    Examiner

    Type

    4.Statistician

    3.Allo

    catio

    nconcealm

    ent

    Reference

    number

    4.Concealment

    adequate

    Serino

    N/A

    1.N/R

    1.N/R

    1.Yes

    1.80%

    1.Yes

    Governm

    ent;

    institu

    tion

    1.N/R

    1.N/R

    High

    2.Patient

    2.N/R

    3.Insufficient

    data

    todeterm

    ine

    3.N/R

    2008

    2.N/R

    2.N/A

    2.No

    2.Unclear

    2.Yes

    CCT

    #48

    3.N/R

    4.N/R

    N/A

    notapplicable;N/R

    notrepo

    rted,Ttest;C

    control;RCTrand

    omised

    controlledtrial;CCTcontrolledclinical

    trial;PRGFplatelet-richgrow

    thfactor;CONSO

    RTCon

    solid

    ated

    Stand

    ards

    ofReportin

    gTrials;ISRCTNInternationalStand

    ardRando

    mised

    Con

    trolledTrial

    Num

    berRegister

    Clin Oral Invest

  • Tab

    le3

    Firstauthor

    Trial

    characteristics

    Population

    characteristics

    Confounding

    factors

    Defect

    characteristics

    Testmaterial

    (num

    berof

    sockets/

    subjects)

    Control

    (num

    berof

    sockets/

    subjects)

    Surgical

    managem

    ent

    Follow-up

    Alveolarridge

    dimension

    changesin

    horizontal

    width

    Alveolarridge

    dimension

    changesin

    vertical

    height

    Implant

    1.Feasibility

    ofim

    plant

    placem

    ent

    2.Necessityof

    simultaneous

    augm

    entation

    Mean/medianmm

    (reference

    point)

    Mean/medianmm

    1.Whole

    ridge

    1.Mid-buccal

    1.Healin

    gperiod

    Year

    ofpublication

    2.Buccalplate

    2.Mesial

    1.Socket

    locatio

    n1.

    Typeof

    flap

    3.Distal

    2.Num

    ber

    ofdrop-outs

    4.SocketFill

    2.Softtissue

    closure

    2.Defect

    morphology

    1.Country

    1.Age

    range

    (mean)

    inyears

    1.Smoking

    3.Adverse

    events

    3.Postoperativ

    eantim

    icrobials

    2.Num

    berof

    centres

    2.Periodontitis

    Type

    Design

    2.Num

    berof

    patients

    (sockets)

    3.Settin

    gMethodology

    Reference

    number

    Aimetti

    1.Italy

    1.36-68

    (51.27

    ±8.4)

    1.No

    1.Maxillary

    anterior

    Calcium

    sulphate

    Empty

    (18/18)

    1.Flapless

    1.3months

    1.T:-2.0±1.1**

    1.T:

    -0.5±1.1*,

    C:-1.2±0.6**,***

    1.Im

    plants

    were

    inserted

    2009

    2.1

    RCT

    Parallel

    2.N/R

    3.University

    Clin

    +Histo

    Hem

    ihydrate

    (22/22)

    #42

    2.40

    (40)

    2.Noprim

    ary

    closure

    2.4-wall

    configuration

    2.N/R

    C:-3.2±1.8**,

    ***

    3.Uneventful

    healing

    3.Amoxicillin

    2g/day

    for

    5days,

    Chlorexidine

    0.12%

    for

    2weeks

    2.N/R

    2.T:-0.2±0.6,

    C:-0.5±0.9

    3.T:-0.4±0.9,

    C:-0.5±1.1

    4.T:

    11.3±2.8**,

    C:10.0±2.3**

    (Acrylic

    stent)

    2.N/R

    Anitua1999

    1.Spain

    1.T:35-55

    (41)

    1.Yes

    1.Any

    T1:

    PRGF

    (5+3/5+3)

    Empty

    (10+3/

    10+3)

    1.Full-thickness

    1.2.5–4

    months

    N/A

    N/A

    1.N/R

    2.N/R

    2.Variable

    2.Yes

    2.Primary

    closure

    2.0

    3.N/R

    T2:

    PRGF+

    Autologous

    bone

    (5/5)

    3.Amoxicillin

    1.5g/day

    for

    5days

    2.1

    C:38-54(42)

    CCT

    3.Private

    practice

    2.23

    (26)

    Parallel+Split-

    mouth

    Histo

    #43

    Barone

    2008

    1.Italy

    1.26-69

    1.<10/day

    1.Non-m

    olars

    Corticocancellous

    porcinebone+

    collagen

    mem

    brane

    (20/20)

    Empty

    (20/20)

    1.Full-thickness

    1.7-9months

    1.T:-2.5±1.2*,

    C:-4.5±0.8*,***

    1.T:-0.7±1.4*,

    C:-3.6±1.5*,***

    1.‘Implantswere

    inserted

    inboth

    groups’

    2.1

    2.Primaryclosure

    2.0

    2.T:-0.2±0.8,

    C:-0.4±1.2

    2.Som

    eGBR

    wereneeded

    dueto

    buccal

    dehiscence

    inthecontrol

    group

    2.N/R

    3.T:-0.4±0.8,

    C:-0.5±1.0

    3.Amoxicillin

    2g/day

    for

    4days+

    Chlorexidine

    0.12%

    for

    3weeks

    3.Uneventful

    healing

    (pain,

    swellin

    g)

    2.Yes

    (treated)

    2.4-wall

    configuration

    4.N/R

    2.40

    (40)

    (Acrylic

    stent)

    3.Hospital

    RCT

    Parallel

    Clin

    +Histo

    #21

    Cam

    argo

    2000

    1.USA,

    Yugoslavia

    1.28-60

    (44±15.9)

    1.N/R

    1.Maxillary

    anterior,

    prem

    olars

    Bioactiv

    eglass+

    coveredby

    calcium

    sulphate

    layer(16/8)

    Empty

    (16/8)

    1.Full-thickness

    with

    4vertical

    releasing

    incisions

    1.6months

    1.T:-3.48

    ±2.68**,

    C:-3.06

    ±2.41**

    1.T:-0.38

    ±3.18,

    C:-1.00

    ±2.25

    (titanium

    tack)

    1.Reentry

    only

    2.N/R

    2.Noprim

    ary

    closure

    3.N/R

    2.N/R

    4.T:

    6.43

    ±2.78**,

    C:4.00±2.33**,***

    (tobuccal

    bone

    crest)

    2.N/A

    3.Penicillin

    1.5g/day

    for

    7days+

    Chlorexidine

    0.12%

    for

    2weeks

    2.N/R

    2.N/R

    2.16

    (32)

    3.Uneventful

    healing

    2.N/R

    CCT

    2.N/R

    3.University

    Split-mouth

    Clin

    #44

    Clin Oral Invest

  • Tab

    le3

    (con

    tinued)

    Firstauthor

    Trial

    characteristics

    Population

    characteristics

    Confounding

    factors

    Defect

    characteristics

    Testmaterial

    (num

    berof

    sockets/

    subjects)

    Control

    (num

    berof

    sockets/

    subjects)

    Surgical

    managem

    ent

    Follow-up

    Alveolarridge

    dimension

    changesin

    horizontal

    width

    Alveolarridge

    dimension

    changesin

    vertical

    height

    Implant

    1.Feasibility

    ofim

    plant

    placem

    ent

    2.Necessity

    ofsimultaneous

    augm

    entation

    Mean/medianmm

    (reference

    point)

    Mean/medianmm

    1.Whole

    ridge

    1.Mid-buccal

    1.Healin

    gperiod

    Year

    ofpublication

    2.Buccalplate

    2.Mesial

    1.Socket

    locatio

    n1.

    Type

    offlap

    3.Distal

    2.Num

    ber

    ofdrop-outs

    4.SocketFill

    2.Softtissue

    closure

    2.Defect

    morphology

    1.Country

    1.Age

    range

    (mean)

    inyears

    1.Smoking

    3.Adverse

    events

    3.Postoperativ

    eantim

    icrobials

    2.Num

    berof

    centres

    2.Periodontitis

    Type

    Design

    2.Num

    berof

    patients

    (sockets)

    3.Settin

    gMethodology

    Reference

    number

    Fiorellini

    2005

    1.USA

    1.47.4

    1.N/R

    1.Maxillary

    anterior,

    prem

    olars

    T1:

    1.5m

    g/ml

    rhBMP-2

    (?/21?)

    Empty

    (?/20?)

    1.Full-thickness

    with

    vertical

    incisions

    1.4months

    1.Coronal:

    T1:+3

    .27±2.53*,

    T2:+1

    .76±1.67*,

    T3:+0

    .82±1.40,

    C:+

    0.57

    ±2.56,***

    (T1vs

    T2/T3/C)

    1.T1:-0.02±1.2,

    T2:-0.62±1.39*,

    T3:-1.00±1.40*,

    C:-1.17±1.23*,

    ***(T1vs

    C)

    1.N/R

    2.Noneed

    for

    augm

    entatio

    n2.

    Nodrop-outs

    reported.

    (3patients

    incorrectly

    random

    i zed,

    1patient

    received

    different

    graft)

    T1:

    18/21(86%

    )T2:

    0.75mg/ml

    rhBMP-2

    (?/22?)

    T2:

    12/22(55%

    )

    2.Primary

    closure

    2.N/R

    T3:

    10/17(59%

    )

    2.N/R

    3.N/R

    C:9/20

    (45%

    )

    3.250(T

    >C)

    4.N/R

    (T1vs

    T2/C)***

    3.Penicillin

    (?mg)

    for7-10

    days+

    Chlorexidine

    0.12%

    T3:

    Collagen

    sponge

    (?/17?)

    2.≥5

    0%buccal

    bone

    loss

    2.8centres

    2.80

    (95)

    2.N/R

    RCT

    3.University

    Parallel

    Radiogr+

    Histo

    #20

    Froum 2002

    1.USA

    1.35-77

    (54.9±11.9)

    1.No

    1.Any

    T1:

    Bioactiv

    eglass(10/8)

    Empty

    (10/10)

    1.Full-thickness

    with

    outvertical

    incisions

    1.6-8months

    N/A

    N/A

    1.‘A

    nim

    plant

    ofappropriate

    size

    was

    placed

    inthehealed

    sockets.’

    2.0

    3.Uneventful

    healing

    T2:

    DFDBA

    (10/8)

    2.Primary

    closure

    2.N/R

    3.Doxycyclin

    e100m

    g/day

    for13

    days+

    Chlorexidine

    0.12%

    for

    30days

    2.N/R

    2.4-wall

    configuration,

    ≤2mm

    buccal

    plateloss

    2.19

    (30)

    2.Single

    centre

    RCT

    Splitmouth

    3.University

    Histo

    #17

    Guarnieri

    2004

    1.Italy

    1.35-58

    1.N/R

    1.Maxillary,

    mandibular

    anteriors,

    prem

    olars

    Calcium

    sulphate

    Empty

    (5/5)

    1.Full-thickness

    with

    outvertical

    incisions

    1.3months

    N/A

    N/A

    1.‘Bucco-lingual

    dimensionsof

    thealveolar

    ridge

    enabled

    safeinsertion

    ofti tanium

    implant.’

    2.socket

    with

    ridge

    resorptio

    n≥5

    0%were

    excluded

    2.Primaryclosure

    3.Amoxicillin

    (?mg)

    for

    1week+Chlorexidine

    0.2%

    for2weeks

    2.N/R

    Hem

    ihydrate

    (10/10)

    2.N/R

    2.Yes

    3.N/R

    2.10

    (25)

    2.N/R

    3.N/R

    CCT

    Parallel+

    Split

    mouth

    Histo

    #45

    Iasella

    2003

    1.USA

    1.28-76

    (51.5±13.6)

    1.Yes

    1.Maxillary

    anteriors,

    prem

    olars

    and

    mandibular

    prem

    olars

    Tetracycline

    hydrated

    FDBA

    +

    Empty

    (12/12)

    1.Full-thickness

    with

    outvertical

    incisions

    1.4or

    6months

    (com

    bined)

    1.T:-1.2±0.9*,

    C:-2.6±2.3*

    1.T:

    +1.3±2.0,

    C:-0.9±1.6***

    1.Im

    plants

    successfully

    placed

    atall

    sites

    2.N/R

    2.N/R

    2.T:

    -0.1±0.7,

    C:-1

    .0±0.8***

    2.Somesiteshad

    slightdehiscence

    andrequired

    furth

    eraugm

    entation

    2.Noprim

    aryclosure

    2.N/R

    3.T:

    -0.1±0.7,

    C:-0.8±0.8***

    3.Doxycyclin

    200m

    g/dayfor

    1week+

    Chlorexidine

    0.12%

    for

    2weeks

    collagen

    mem

    brane

    (12/12)

    2.0

    4.N/R

    3.N/R

    (Acrylic

    stent)

    3.N/R

    2.24

    (24)

    RCT

    2.N/R

    Parallel

    Clin

    +Histo

    #23

    Clin Oral Invest

  • Tab

    le3

    (con

    tinued)

    Firstauthor

    Trial

    characteristics

    Population

    characteristics

    Confounding

    factors

    Defect

    characteristics

    Testmaterial

    (num

    berof

    sockets/

    subjects)

    Control

    (num

    berof

    sockets/

    subjects)

    Surgical

    managem

    ent

    Follow-up

    Alveolarridge

    dimension

    changesin

    horizontal

    width

    Alveolarridge

    dimension

    changesin

    vertical

    height

    Implant

    1.Feasibility

    ofim

    plant

    placem

    ent

    2.Necessity

    ofsimultaneous

    augm

    entation

    Mean/medianmm

    (reference

    point)

    Mean/medianmm

    1.Whole

    ridge

    1.Mid-buccal

    1.Healin

    gperiod

    Year

    ofpublication

    2.Buccalplate

    2.Mesial

    1.Socket

    locatio

    n1.

    Type

    offlap

    3.Distal

    2.Num

    ber

    ofdrop-outs

    4.SocketFill

    2.Softtissue

    closure

    2.Defect

    morphology

    1.Country

    1.Age

    range

    (mean)

    inyears

    1.Smoking

    3.Adverse

    events

    3.Postoperativ

    eantim

    icrobials

    2.Num

    berof

    centres

    2.Periodontitis

    Type

    Design

    2.Num

    berof

    patients

    (sockets)

    3.Settin

    gMethodology

    Reference

    number

    Lekovic

    1997

    1.Yugoslavia

    /USA

    1.(49.8)

    1.N/R

    1.Maxillary

    and

    mandibular

    anteriors,

    prem

    olars

    e-PTFE

    mem

    brane

    (10/10)

    Empty

    (10/10)

    1.Full-thickness

    with

    4vertical

    releasing

    incisions

    1.6months

    1.10/10:,

    T:-1.80±0.51,

    C:-4.40±0.61**,

    ***

    1.10/10:

    T:-0.5±0.22,

    C:-1.2±0.13**,

    ***

    1.Reentry

    only

    7/10:T:-1.71±0.75,

    C:-4.43±0.72**,

    ***

    7/10:

    T:-0.28±0.18,

    C:-1.0±0.0**,

    ***

    3/10:T:-2.00±0.00,

    C:-4.33±0.88*

    3/10:T:-1.0±0.58,

    C:-1.66±0.33

    (titanium

    tack)

    2.N/R

    3.N/R

    4.10/10:

    T:4.9±0.86*,

    C:3.0±0.63,

    ***

    2.N/R

    7/10:T

    :5.43±1.1*,

    C:2.92±1.61,***

    3/10:T

    :3.66±1.20,

    C:4

    .33±1.45

    (tobuccalbone

    crest)

    2.Primaryclosure

    3.Penicillin

    1g/day

    for

    7days+

    Chlorexidine

    0.2%

    2.N/A

    2.3/10

    drop-outs

    dueto

    prem

    ature

    mem

    brane

    exposure

    3.3/10

    exposed,

    7/10

    noinfection

    2.10

    (20)

    2.N/R

    CCT

    2.N/R

    (presumably

    singlecentre)

    Split-mouth

    2.N/R

    Clin

    3.University

    #24

    Lekovic

    1998

    1.Yugoslavia

    1.(52.6±11.8)

    1.N/R

    1.Maxillary

    and

    mandibular

    anteriors,

    prem

    olars

    PG/PL

    mem

    brane

    (16/16)

    Empty

    (16/16)

    1.Full-thickness

    with

    4vertical

    releasingincisions

    1.6months

    1.T:-1.31±0.24*

    1.T:-0.38

    ±0.22,

    C:-1.50

    ±0.26*,

    ***(titanium

    tack)

    1.Reentry

    only

    2.N/A

    2.0

    C:-4.56

    ±0.33*,

    ***

    3.Uneventful

    healing

    2.N/R

    2.Primaryclosure

    2.N/R

    3.Penicillin1g/day

    for7days+

    Chlorexidine0.12%

    for2weeks

    3.N/R

    2.Yes

    (treated)

    2.N/R

    4.T:5.81

    ±0.29*,

    C:3.94

    ±0.35*,

    ***

    (tobuccal

    bone

    crest)

    2.16

    (32)

    2.1

    3.University

    RCT

    Split-mouth

    Clin

    #25

    Nevins2006

    1.USA

    /Italy

    1.N/R

    1.N/R

    1.Maxillaryanterior

    DBBM

    (19/9)

    Empty(17/9)

    1.Partial

    thickness

    1.1–3months

    (biopsiesat6M

    )N/A

    1.T:-2.42±2.58,

    C:-5.24±3.72***

    1.Im

    plantswere

    placed,but

    number

    unknow

    n

    2.Buccalplatewas

    comprom

    ised

    2.Primary

    closure

    2.0

    3.N/R

    2.N/R

    2.N/R

    3.N/R

    3.N/R

    4.N/R

    (At6mm

    ridgewidth)

    2.Yes

    RCT

    2.N/R

    Split-mouth

    3.N/R

    Radiogr+

    Histo

    #46

    2.9(36)

    Clin Oral Invest

  • Tab

    le3

    (con

    tinued)

    Firstauthor

    Trial

    characteristics

    Population

    characteristics

    Confounding

    factors

    Defect

    characteristics

    Testmaterial

    (num

    berof

    sockets/

    subjects)

    Control

    (num

    berof

    sockets/

    subjects)

    Surgical

    managem

    ent

    Follow-up

    Alveolarridge

    dimension

    changesin

    horizontal

    width

    Alveolarridge

    dimension

    changesin

    vertical

    height

    Implant

    1.Feasibility

    ofim

    plant

    placem

    ent

    2.Necessity

    ofsimultaneous

    augm

    entation

    Mean/medianmm

    (reference

    point)

    Mean/medianmm

    1.Whole

    ridge

    1.Mid-buccal

    1.Healin

    gperiod

    Year

    ofpublication

    2.Buccalplate

    2.Mesial

    1.Socket

    locatio

    n1.

    Type

    offlap

    3.Distal

    2.Num

    ber

    ofdrop-outs

    4.SocketFill

    2.Softtissue

    closure

    2.Defect

    morphology

    1.Country

    1.Age

    range

    (mean)

    inyears

    1.Smoking

    3.Adverse

    events

    3.Postoperativ

    eantim

    icrobials

    2.Num

    berof

    centres

    2.Periodontitis

    Type

    Design

    2.Num

    berof

    patients

    (sockets)

    3.Settin

    gMethodology

    Reference

    number

    Pelegrine

    2010

    1.Brazil

    1.28-70

    (47.5±10.3)

    1.No

    1.Maxillary

    anteriors

    Autologous

    bone

    marrow

    (15/7)

    Empty

    (15/6)

    1.Full-thickness

    with

    2buccal

    vertical

    releasing

    incisions

    1.6months

    1.T:-1.0*,

    C:-2.5*,***

    1.T:

    -0.5*,

    C:-1.0*,***

    (Titanium

    screw)

    1.Allimplants

    osseointegrated

    2.N/R

    2.13

    (30)

    2.N/R

    2.Primaryclosure

    3.N/R

    2.1

    2.0

    3.University

    2.T:-0.75,

    C:-1.75,***

    4.T:10.33*,C:10.32*

    (tobuccal

    bone

    crest)

    2.T:with

    out

    further

    augm

    entatio

    n,C:At5sites

    augm

    entatio

    nor

    expansion

    carriedout

    3.N/R

    3.Uneventful

    healing

    RCT

    2.Socketswith

    severe

    bone

    loss

    were

    excluded

    Parallel

    Clin

    +histo

    #47

    Serino

    2003

    1.Italy

    1.35-64

    1.N/R

    1.Any

    PG/PLsponge

    (26/24)after

    drop-out

    Empty

    (13/12)

    afterdrop-out

    1.Full-thickness

    buccally

    and

    lingually

    1.6months

    N/A

    1.T:+1.3±1.9*,

    C:-0.8±1.6

    1.Placementof

    implantsin

    all

    CandTsites

    with

    good

    prim

    arystability

    2.Buccalplate

    couldbe

    partially

    orcompletely

    lost

    2.1

    CCT

    Parallel+

    split-m

    outh

    2.45

    (39)

    before

    drop-out

    3.N/R

    2.Yes

    (treated)

    Clin

    +Histo

    #19

    2.Noprim

    ary

    closure

    2.9drop-outs

    forreasons

    unrelatedto

    thetherapy

    2.T:-0.2±1.0,

    C:-0.6±1.0

    3.T:-0.1±1.1,

    C:-0.8±1.5

    3.No

    antib

    iotics;

    Chlorexidine

    0.2%

    for

    2weeks

    3.Uneventful

    healing

    4.N/R

    (Acrylic

    stent)

    2.N/R

    Serino

    2008

    1.Italy

    1.32-64

    1.N/R

    1.Any

    non-molars

    PG/PLsponge

    (7/7)

    Empty(9/9)

    afterdrop-out

    1.Full-thickness

    buccally

    and

    lingually

    1.3months

    N/A

    N/A

    1.Placementof

    implantsin

    all

    CandTsites

    with

    good

    prim

    arystability

    2.1

    2.Yes

    (treated)

    2.4drop-outs

    forreasons

    unrelatedto

    thetherapy

    afterdrop-out

    CCT

    2.Alveolar

    bone

    height

    ≥8mm

    2.Noprim

    ary

    closure

    Parallel

    2.N/R

    Histo

    3.NoAntibiotics;

    Chlorexidine

    0.2%

    for2weeks

    3.Uneventful

    healing

    2.20

    (20)

    before

    drop-out

    #48

    3.N/R

    *p<0.05;statistically

    sign

    ificantintra-grou

    pdifference,baselin

    eto

    final;**

    p<0.00

    1statistically

    high

    lysign

    ificantintra-grou

    pdifference,baselin

    eto

    final;**

    *p<0.05

    statistically

    sign

    ificant

    inter-groupdifference,betweentestandcontrol;

    N/A

    notapplicable;N/R

    notrepo

    rted;Ttest;C

    control;M=mon

    th(s);Clin

    clinical

    analysis;Histo

    histolog

    ical

    analysis;Rad

    iogr

    radiog

    raph

    icanalysis;RCTrand

    omised

    controlledtrial;CCT

    controlledclinicaltrial;PRGFplasmarich

    ingrow

    thfactors;DFDBAdemineralised

    freeze-dried

    bone

    allograft;FDBAfreeze-dried

    bone

    allograft;e-PTFEexpand

    ed-polytetrafluo

    rethylen;PG/PL

    polyglycolide/po

    lylactide;

    DBBM

    demineralised

    bovine-bon

    emineral

    Clin Oral Invest

  • Tab

    le4

    Firstauthor

    Num

    berof

    biopsies

    Histomorphology

    Histomorphom

    etry

    Statistical

    difference

    betweentestand

    control

    Test

    Control

    (meanor

    median%)

    Yearof

    publication

    (testmaterial)

    Type

    Healin

    gperiod

    Reference

    number

    Aimetti2009

    T:N/R

    22?

    (MGCSH)

    Noresidual

    graftmaterial.Noinflam

    matory

    infiltrate.New

    bone

    form

    ationin

    all

    specim

    ens,100%

    livingtrabecular

    bone

    with

    woven

    andlamellarstructure.

    100%

    livingbone

    (mostly

    woven)

    inallbiopsies.Lam

    ellarbone

    remodelingwas

    starting.

    Trabecular

    bone:

    Residual

    substitute

    material:

    Woven

    bone:

    Lam

    ellar

    bone:

    Tvs

    C*

    T:58.8±3.5

    Coronal:

    Coronal:

    T:0.0

    T:83.6±6.6

    T:16.4±6.6

    C:11.1±7.6

    C:88.9±7.6

    Middle:

    T:40.4±13.2

    Middle:

    T:59.6±13.2

    C:18.9±7.6

    C:81.1±7.6

    Apical:

    Apical:

    T:56.4±10.9

    T:43.6±10.9

    C:77.8±8.1

    RCT

    C:N/R

    18?

    C:22.2±8.1

    C:N/A

    C:47.2±7.7

    3M

    #42

    Anitua

    1999

    T:N/R

    (PRGF±

    autogenbone)

    Com

    pact

    maturebone

    with

    well-organized

    trabeculae

    andmorphologyin

    8/10

    patients.

    Connectivetissuewith

    non-organized

    trabeculae

    in2/10

    patients.Significant

    intra-groupdifferences10

    vs.16

    weeks!

    Connectivetissuefills

    themain

    partof

    thedefect.Nomature

    bone.

    N/R

    CCT

    2.5–4M

    C:N/R

    #43

    Barone2008

    T:20 (Corticocancellous

    porcinebone+

    collagen

    mem

    brane)

    Residualgraftmaterialem

    bedded

    innewly

    form

    edbone

    inallspecim

    ens.Com

    plete

    bone

    fill.

    Typically

    trabecular

    bone

    pattern.

    Large

    marrow

    spaces

    filledwith

    adipocytes.Lam

    ellarbone

    was

    also

    presentwith

    inthebone

    marrow.

    Totalbone

    volume:

    Connective

    tissue:

    Residualgraft

    material:

    Bone:

    T>C*

    Connectivetissue:

    T<C*

    RCT

    T:35.5±10.4

    T:36.6±12.6

    T:29.2±10.1

    C:25.7±9.5

    C:59.1±10.4

    C:N/A

    7–9M

    C:20

    #21

    Fiorellini

    2005

    T1:16

    (rhB

    MP-2

    1.5m

    g/ml)

    Noevidence

    ofinflam

    mationor

    residualgraft.Trabecularbone

    form

    ationin

    2/3

    ofthesamples.M

    ineralized

    tissueform

    ationpresentedwith

    differentlevelof

    remodeling.Minor

    osteoclasticactiv

    ity.N

    ocomparisonreported

    between

    TandC!

    N/R

    T2:

    15(0,75m

    g/ml)

    T3:11

    (Collagensponge)

    RCT

    C:14

    4M

    #20

    Froum

    2002

    T1:

    10(Bioactiv

    eglass)

    T1:

    New

    bone

    form

    ation.

    Osteoid

    surrounded

    andpenetrated

    thebioactiveglassparticles.

    N/R

    Vitalbone:

    Connective

    tissue:

    Residualbone

    substitute:

    Connectivetissue:

    T1<T2or

    C*

    T1:

    59.5

    T1:

    35.3

    T1:

    5.5

    T2:

    13.5

    T2:

    34.7

    T2:

    51.6

    C:N/A

    C:67.0

    C:32.4

    T2:

    Varying

    degreesof

    reossificatio

    naround

    DFDBA.

    RCT

    T2:

    10(D

    FDBA)

    C:10

    6–8M

    #17

    Clin Oral Invest

  • Tab

    le4

    (con

    tinued)

    Firstauthor

    Num

    berof

    biopsies

    Histomorphology

    Histomorphom

    etry

    Statistical

    difference

    betweentestand

    control

    Test

    Control

    (meanor

    median%)

    Yearof

    publication

    (testmaterial)

    Type

    Healin

    gperiod

    Reference

    number

    Guarnieri2004

    T:10

    (MGCSH)

    Alm

    ostcompleteabsenceof

    MGCSH.

    Absence

    ofconnectiv

    etissueand

    inflam

    matorycells.In

    allsections

    trabecular

    bone

    form

    ationwith

    nodifferencesbetweentheapical,middle

    andcoronallevels.

    Lessbone

    form

    ationcompared

    totestsites.

    Trabecular

    bone

    area:

    Nostatisticalsignificance

    couldbe

    draw

    ndue

    tosm

    allnum

    berof

    controlspecimens.

    CCT

    C:5

    T:Coronal:58.6±9.2

    3M

    Middle:

    58.1±6.2

    Apical:58.3±7.8

    C:≤46

    #45

    Iasella

    2003

    T:4M

    :5,

    6M:7

    Residualgraftparticlessurrounded

    bywoven

    bone

    orby

    connectiv

    etissue.

    Similaram

    ount

    oftotalb

    oneand

    trabecular

    spaces

    asin

    test.

    Vitalbone:

    Non-vitalbone:

    N/R

    (Tetracycline

    hydrated

    FDBA+

    Collagen

    mem

    brane)

    (Nobiopsy

    from

    2Csitesdue

    tominim

    albone

    fill)

    4M4M

    T:31

    ±9

    T:32

    ±19

    C:58

    ±11

    C:N/A

    6M T:41

    ±18

    C:N/A

    Com

    bined

    6M

    T:37

    ±18

    C:N/A

    T:25

    ±17

    C:50

    ±14

    Com

    bined

    RCT

    T:28

    ±14

    C:54

    ±12

    4–6M

    #23

    C:4M

    :5,

    6M:5

    Nevins2005

    T:5(D

    BBM)

    DBBM

    granules

    present.Apically

    integrated

    incancellous

    bone

    butcoronally

    insoft

    tissue.Nosignsof

    inflam

    mationor

    foreignbody

    reactio

    n.

    New

    bone

    form

    ation

    Nocomparisonmade.

    RCT

    6M

    #46

    C:5

    Pelegrine

    2010

    T:7

    (Autologous

    bone

    marrow)

    Mineralized

    bone:

    Nosignificantdifference.

    T:45.0

    RCT

    C:43.75

    C:6

    6M

    #47

    Serino

    2003

    T:10 (PG/PLsponge)

    Noresidual

    graftmaterial.Presenceof

    matured,mineralized

    bone.Lackof

    coronalsofttissueingrow

    th.

    Presenceof

    mineralized

    bone.

    Widemarrow

    spaces.

    Mineralized

    bone:

    Statisticalcomparison

    cannotbe

    madedue

    tothesm

    allnum

    ber

    ofcontrolspecimens.

    T:66.7

    C:3

    C:43.7

    CCT

    6M

    #19

    Serino

    2008

    T:7 (PG/PLsponge)

    Noresidualgraftm

    aterial.Scarce

    presence

    ofinfl ammatorytissue.Coronal:new

    lyform

    edtrabecularbone

    with

    largemarrowspaces.

    Apical:morematureandcompactbone.

    Coronal:trabecularbone

    with

    wide

    marrowspaces

    with

    connective

    tissue.Apical:morematureand

    compactbone.

    Mineralized

    bone:

    Nosignificant

    difference.

    T:59.9±22.4

    CCT

    C:9

    C:48.8±14.4

    3M

    #48

    *p<0.05

    ;statistically

    sign

    ificantdifference

    betweentestandcontrol

    Ttest;Ccontrol;M

    mon

    th(s);N/R

    notrepo

    rted;N/A

    notapplicable;vs.v

    ersus;TBVtotalbo

    nevo

    lume;MGCSH

    medicalgradecalcium

    sulphatehemihyd

    rate;DFDBAdemineralised

    freeze-dried

    bone

    allograft;FDBAmineralised

    freeze-dried

    bone

    allograft;DBBM

    demineralised

    bovine-bon

    emineral;PG/PLpo

    lyglycolide/po

    lylactide

    Clin Oral Invest

  • prerequisite of this technique would be some type of stand-ardisation, so that the captured image is being always takenfrom exactly identical positions [54]. None of the two includedradiographic studies reported on such standardisation [20, 46].

    For the interpretation of the results we attempted tocluster the studies in respect to the type of intervention.

    Unassisted sockets In the present review, the mean reduc-tion of the AR width of the untreated sites varied between2.6±2.3 mm and 4.6±0.3 mm and the mean reduction of theAR height was between 0.8±1.6 mm and 3.6±1.5 mm after1 to 9 months of healing. This corroborates the result of aprevious clinical study which indicated that 95 % of ARreduction should be expected after three months of extrac-tion [1]. Furthermore, it is in agreement with a recent sys-tematic review, which reported that the average reduction ofthe AR width seemed to be higher (3.87 mm), than thereduction in AR height (1.67 mm) [55].

    Even though both AR width and height present resorp-tion, histologically, new bone formation up to a variableextent was also observed in some studies as result of unas-sisted socket healing [19–21, 23, 42, 45, 46, 48]. In addi-tion, a large area was occupied by bone marrow [19, 21, 48],as reported in preclinical studies [11, 13, 56]. Only a singlestudy reported on connective tissue fill and lack of maturebone [43].

    Bone grafts and substitutes Effective grafting procedures forbone augmentation have been associated with the osteocon-ductive, osteoinductive or osteogenetic properties of the graft[56–59]. This led to the assumption that the placement of thesematerials in the extraction socket may accelerate new boneformation by the above biological properties and may alsoreduce AR resorption by stabilising the blood clot, providing ascaffold and external source of minerals and/or collagen [11,12, 60, 61]. The placement of DBBM with collagen in freshextraction sockets resulted in limited reduction of the ARdimensions, although delayed initial socket healing in termsof new bone formation was also observed [11, 12]. Humanstudies reported similar unfavourable histological observa-tions when DFDBAwas employed for ARP [15, 16].

    In the present review of human experiments, two out ofthree studies reported that socket grafting with autologousbone marrow [47] or alloplastic material [42] have signifi-cantly limited the reduction of the AR width compared tothe unassisted socket healing. Three out of five studiesreported that reduction of the resorption in AR height wassignificant [42, 46, 47], while the ridge height was evenincreased in one study, where sockets were grafted withpolymer sponge [19]. We should emphasise though thatsince the graft material (DBBM) in a CT study possessedradiopaque characteristic, the alteration of the AR contouron the CT image should be interpreted with caution [46].

    Based on the histological evaluation of these studies, theabove AR dimensional changes were not necessarily accom-panied by higher amount of new bone formation in thesocket, since the quality of newly formed tissue in theARP sites was comparable to that in the control sites.Furthermore, the sockets were occupied by a mixture ofnew bone and connective tissue which in many occasionswas surrounding the graft particles [17, 21, 46] (Table 4).

    GBR (membrane alone or in combination with ‘graft’) Theconception of guided bone and tissue regeneration [62] wastranslated to ARP procedures in order to exclude epithelialcells from the extraction socket by the use of barrier mem-brane in four studies of the present review [21, 23–25].

    (a) GBR with membrane aloneARP with GBR resulted in statistically significantly

    less resorption in ridge width and height compared tounassisted socket healing, regardless of the type ofmembrane [24, 25]. It should be noted that in one study[24], in three out of 10 cases, the exposed non-resorbable e-PTFE barrier had to be removed prema-turely, highlighting the importance of sufficient softtissue closure and timing of removal of the barrier.The outcomes in these three cases were similar to thecontrol sites. Where healing was uncompromised, astatistically significant difference was found after6 months in width and height changes in favour ofthe ARP group.

    (b) GBR with membrane and ‘graft’ARP resulted in statistically significantly less re-

    sorption in width [21, 23] and height [23] in compar-ison to unassisted socket healing. The histologicalevaluation of the GBR procedures in the includedstudies demonstrated new bone formation [21, 23],but the presence of graft particles was also evident inboth studies, embedded either in newly formed bone[21] or in connective tissue [23]. This is in agreementwith a recent trial, where a collagen membrane incombination with DBBM or a biphasic bone substitutewas used for ARP [9, 10].

    Biological active materials The potential benefit of biologi-cal active molecules was investigated in periodontal and boneregeneration through fostering the proliferation and differen-tiation of different mesenchymal cells in various preclinicalmodels [63, 64]. The safety and feasibility of rhBMP-2 onhuman ARP or ridge augmentation was evaluated and shownto be safe in a two-centre clinical study [35]. Dimensionalchanges of the alveolar ridge were measured on CTscans in anRCT [20]. Treatment with recombinant BMP-2 resulted in anincrease in ridge width which was statistically significantlygreater than controls. However, this observation needs to be

    Clin Oral Invest

  • interpreted in light of the surprise finding of an increase inridge width of the untreated controls. This was a uniquefinding amongst the studies that we reviewed. Histologically,no comparison between ARP and controls sites was reported.

    The human histological results of the included papers ofthe present review were generally found to be comparable topreclinical studies [11–13, 60, 65]. There are a number ofaspects to consider in the interpretation of the results. First-ly, it has to be kept in mind that whilst the biopsies of theanimal model incorporate the cross section of the whole AR,the biopsy retrieval at human studies is limited to a trephinecore sample of part of the former socket. This location maynot necessarily coincide with the exact position of the pre-vious extraction, thus making interpretation of the resultschallenging. Furthermore, the differentiation between api-cal, mid and coronal, as well as the central and lateralaspects of the biopsies was not always apparent.

    Another important parameter when considering a histo-logical overview of the studies was the variation in healingtime. Due to the nature of post-extraction healing, the directcomparison of the new tissue formation in studies between 1and 9 months of healing could be misleading. This washighlighted in three studies which did not make a distinctionbetween the variable healing times within the groups, rang-ing from 2.5 months to 9 months [17, 21, 43]. It has to bekept in mind also that the only study, which completed andreported appropriate statistical methodology [47], did notobserve statistically significant difference between the testand control biopsies.

    Furthermore, small sample sizes in the majority of thestudies may also limit the generalisability of the histologicalfindings.

    Two studies found statistically significant histologicaldifferences in new bone formation favouring the test group[21, 42]. Drawing conclusions across the studies is difficultsince the test groups differed in many respects comparedwith each other, including different technique (bone sub-stitute only [42]/GBR + graft [21]), different material(MGCSH [42]/porcine bone with collagen membrane[21]), different flap management (flapless, no primary clo-sure [42]/mucoperiosteal flap, primary closure [21]), dif-ferent healing time (3 months [42]/7–9 months [21]). Onecommon feature was that both groups limited their inter-vention to sockets with four intact walls. It is noteworthythat all three studies that included intact socket walls only,reported statistically significant differences both on ARwidth and height in favour of ARP [21, 42, 47], while onlyone [20] out of two studies [19, 20] with initial buccal boneloss reported similar significant difference between testand control. Therefore, socket morphology could be animportant predicator of improved ARP. The need forARP in such sockets, in terms of future clinical success/implant placement needs further investigation.

    Flap management All studies reporting statistically signifi-cant inter-group differences in both horizontal and verticalclinical measurements achieved either primary flap closure[21, 24, 25, 47], or did not detach the periosteum in aflapless procedure [42]. Furthermore, none of the studieswithout primary closure demonstrated statistically signifi-cant differences between test and control in terms of bothhorizontal and vertical clinical measurements [19, 23, 44].Therefore, both achieving and maintaining the epithelialseal above the socket may be crucial to improving ARP.Further corroboration of this concept was suggested wheree-PTFE barriers were prematurely exposed. The healing ofthese three exposed cases demonstrated no statistically sig-nificant differences compared to the control sites [24].

    Other factors affecting interpretation of the findings

    Healing time

    The optimal timing of re-entry following ARP is determinedby the implant insertion. Since the volume of the AR isgradually decreasing, while the quality of the newly formedtissue is gradually increasing during the post-extractionremodelling [1, 6] the implant placement could be consid-ered as early as possible, but as late as necessary, in order tomaintain AR volume, as well as to achieve complete epi-thelial seal with some extent of osseous fill. The healingperiods of the trials in the present review varied consider-ably (one to nine months). Therefore, interpretation of theresults was complicated by the heterogeneity present in theincluded studies.

    Antimicrobials

    Improvement of clinical parameters was demonstrated as aresult of regular rinsing with chlorhexidine following toothextraction [66]. Subjects of the included trials in the presentreview were prescribed various types of antibiotics andinstructed to rinse with chlorhexidine for 2 to 3 weeks.Therefore, no conclusion could be drawn on the necessityor benefit of employment of antibiotics/antimicrobials fol-lowing ARP.

    Smoking

    Smoking is associated with delayed socket healing andincreased reduction in post-extraction alveolar width [67].Three trials in this review included smokers [21, 23, 43] andthe half of the studies did not report on smoking as anexclusion factor, thus any conclusions about the impact ofthis well-recognised risk factor for impaired healing aredifficult to draw [68].

    Clin Oral Invest

  • Periodontal treatment/health

    Four studies included patients whose periodontal treatmentwas carried out prior to the ARP [19, 21, 25, 48]. ARPresulted in statistically significant difference between testsand controls in clinical [21, 25] and in histological param-eters [21]. In addition, in the studies where periodontitis waspresent, but periodontal treatment was not reported, nostatistically significant histological differences were demon-strated [43, 44, 46]. This suggests that treated periodontitismay not hinder the success of ARP.

    Hard and soft tissue morphology

    No data were reported on factors, such as gingival biotype,width of the keratinised gingiva, thickness of buccal plate ortotal volume of AR that may modify the outcome of ARP.Therefore, the possible impact of these factors on ARPcannot be determined.

    Clinical relevance

    The clinical rationale for ARP is to minimise the necessityfor one or two stage alveolar ridge reconstruction to allowsuccessful implant placement. If the ARP procedure failsto meet this requirement, it may be considered as an un-necessary or even unsuccessful procedure. Therefore, astatistical significance favouring ARP does not necessarilylead to a clinical benefit, unless the whole treatment issimplified or made more successful [9]. In the presentsystematic review, seven out of ten studies did not reportdifferences in feasibility of implant insertion at re-entry[17, 19, 23, 42, 45, 46, 48]. Only two studies reported thatthere was no need for further reconstruction in the ARPgroup, whilst GBR or ridge expansion were carried out insome of the control sites alongside implant insertion [21,47]. One study reported that statistically significantly lessaugmentation had to be performed in the ARP group,compared to the control [20]. In relation to illuminatingthe understanding of possible long term benefits of ARP,the success rate of the inserted dental implants in theformer test, versus control sites should be examined. Nostudies have yet reported this.

    Patient-reported outcome and health economics

    It would be helpful to understand patient experiences suchas concomitant discomfort at/following ARP in order toavoid a further, extensive reconstructive surgery. On theother hand, the additional costs of ARP at the time ofextraction may not be desirable if the outcome and benefitof such extra treatment were not predictable. There are nodata yet to inform on these questions.

    Conclusions

    Within the limits of the above findings the following con-clusions can be drawn:

    1. The results of the control groups confirm that toothextraction results in a statistically significant horizontaland vertical resorption of the AR, as part of the naturalremodelling.

    2. The magnitude of the horizontal shrinkage is morepronounced than the vertical.

    3. The resorption of the AR cannot be totally preventedby ARP.

    4. Dimensional changes of the AR may be limited bysome of the ARP techniques.

    5. No evidence was identified to inform on the possibleimpact of the following factors on ARP outcomes: (a)site location, (b) buccal plate thickness, (c) healingtime, (d) antibiotic regime, (e) light smoking, (f) his-tory of treated periodontitis.

    6. The presence of intact socket walls and primary flapclosure are often associated with favourable results.

    7. Conflicting evidence exists on the benefit of ARPat the histological level. ARP does not appear topromote de novo hard tissue formation routinely.In addition, some graft materials may interfere withhealing.

    8. Due to the broad variety of employed materials, tech-niques, defect morphologies, healing periods, as wellas the relatively small sample sizes, meta-analysis orcomparative assessment of ARP cannot be made. Con-sequently no material or method can be claimed toserve superior to another. However, in certain casesGBR appeared to be most effective.

    9. Only limited evidence supports the clinical benefit ofARP, namely the reduction of necessity of furtheraugmentation in conjunction with implant placement.

    10. No evidence exists on comparison of the survival orsuccess rate of implants, placed in the former ARP orcontrol sites.

    11. No evidence exists on cost-effectiveness, patient’spreference or quality of life following ARP.

    12. The case selection criteria for performing ARP remainstill undetermined.

    13. The strength of evidence ranges from weak to moder-ate and therefore, the conclusions of this review shouldbe interpreted with caution.

    Recommendations for further research

    & Randomised controlled trials on adequately poweredsample sizes are needed where unassisted socket healingserves as the negative control.

    Clin Oral Invest

  • & Appropriate follow-up periods are required. Ideally, thisshould reflect implant insertion protocols, such as sixweeks (Type 2), three to four months (Type 3) or>6 months (Type 4) placement following extraction.

    & Clinical studies should be designed to perform not onlyclinical (quantitative), but also histological (qualitative)assessment.

    & The role of additional factors like smoking, reason forextraction, tooth location, initial buccal plate thickness,flap reflection and closure, antimicrobial regime shouldalso be investigated.

    & Comparative studies should also be designed in order toidentify the most successful treatment options.

    & It may be beneficial to seek for a cell occlusive barriermembrane that does not require extensive soft tissuemobilization for flap approximation.

    & Necessity of re-augmentation at implant placementshould be investigated.

    & Survival and success rates of implants, placed in formerARP sites should be evaluated.

    & Outcome evaluation should ideally incorporate patient’spreference, quality of life, as well as treatment economy.

    Acknowledgments The authors wish to express their gratitude toAviva Petrie for her invaluable contribution to the statistical assessment.

    Conflict of interest and source of funding There was no known conflictof interest among the review team. The trial was self funded and supportedby the Research Discretionary Account of the Unit of Periodontology, UCLEastman Dental Institute. This work was undertaken at UCLH/UCL whoreceived a proportion of funding from the Department of Health’s NIHRBiomedical Research Centres funding scheme.

    References

    1. Schropp L, Wenzel A, Kostopoulos L, Karring T (2003) Bonehealing and soft tissue contour changes following single-toothextraction: a clinical and radiographic 12-month prospective study.Int J Periodontics Restor Dent 23:313–323

    2. Amler MH, Johnson PL, Salman I (1960) Histological and histo-chemical investigation of human alveolar socket healing in undis-turbed extraction wounds. J Am Dent Assoc 61:47–58

    3. AmlerMH (1969) The time sequence of tissue regeneration in humanextraction wounds. Oral Surg Oral Med Oral Pathol 27:309–318

    4. Araújo MG, Lindhe J (2005) Dimensional ridge alterations follow-ing tooth extraction. An experimental study in the dog. J ClinPeriodontol 32:212–218

    5. Boyne PJ (1966) Osseous repair of the postextraction alveolus inman. Oral Surg Oral Med Oral Pathol 21:805–813

    6. Cardaropoli G, Araújo M, Lindhe J (2003) Dynamics of bonetissue formation in tooth extraction sites. An experimental studyin dogs. J Clin Periodontol 30:809–818

    7. Claflin RS (1936) Healing of disturbed and undisturbed e