The Effectiveness of Surgical Management of Oroantral Communications: a Systematic Review of the Literature

SYSTEMATIC REVIEW 347

Matteo De Biasi, Michele Maglione, Daniele Angerame The effectiveness of surgical management of oroantral communications: A systematic review of the literature

Matteo De Biasi, DDS, MS PhD Student, Graduate School of Nanotechnology, Key words extraction socket, graft, oroantral communication, oroantral ﬁstula University of Trieste, Trieste, Italy

Background: An oroantral communication (OAC) is a common complication in alveolar surgery that Michele Maglione, MD, DDS usually occurs as a result of the extraction of maxillary posterior teeth. To avoid further complications, Professor, Dental Clinic, several closure techniques are used; most of them need a flap elevation. Recently, simpler conserva- University Clinic Department of Medical, Surgical and tive flapless techniques for OAC closure have been described. Health Sciences, University Objectives: To appraise the effectiveness of different techniques for closure of OACs also in com- of Trieste, Trieste, Italy parison to nothing. Daniele Angerame, MD, DDS Search methods: The following electronic databases were searched for randomised controlled trials Professor, Dental Clinic, regarding techniques for closure of OACs: PubMed; SciVerse Scopus; Latin American and Caribbean University Clinic Department of Medical, Surgical and Health Sciences; The Scientific Electronic Library Online and The Cochrane Library (from January Health Sciences, University 1949 to August 2014). Unspecific algorithms were chosen in order to maximise search sensibility. of Trieste, Trieste, Italy Additional manual searching was performed in PubMed related citations, in five journals and in the Correspondence to: references of the selected articles. There were no restrictions with regard to publication language. Daniele Angerame, Dental Clinic, Selection criteria: Randomised controlled trials (RCTs) comparing techniques for closing oroantral University Clinic Department of Medical, communications to nothing, or different techniques for closing oroantral communications reporting Surgical and Health the success rate with at least two months follow-up. Sciences, University of Trieste, Data collection and analysis: The screening of eligible studies, the assessment of methodological Piazza Ospedale 1, 34129 quality and data extraction were done by two independent reviewers working in duplicate. Trieste, Italy Results: The research individuated 1256 publications. After screening, only five articles were assessed Tel: +39 0403992761 for eligibility. Only two RCTs evaluating the effectiveness of techniques for OAC closure fulfilled the Fax: +390403992665 Email: d.angerame@fmc. inclusion criteria of the present review. One trial including 30 patients assessed whether flapless tech- units.it niques (resorbable root analogues and haemostatic gauze) could be as effective as the Rehrmann’s buccal flap; all the patients were reported as successfully healed in the three intervention groups. Another RCT with 20 patients compared the effectiveness of the buccal fat pad flap (100% success rate) with a sandwich graft with hydroxyapatite crystals within collagen sheaths (90% success rate). The authors found no significant difference. Conclusions: There are no RCTs evaluating whether an oroantral communication should be closed or not. There is weak evidence from two RCTs showing good results with five different techniques for closure of OACs (resorbable root analogues, haemostatic gauze, Rehrmann’s buccal flap, buccal fat pad flap, sandwich graft with hydroxyapatite crystals). Until sufficiently high quality RCTs are conducted, elevating or not a flap for closure of OACs will be left to the personal choice of the surgeon.

Conflict of interest notification: The authors declare no competing financial interest.

Eur J Oral Implantol 2014;7(4):347–357 348 De Biasi et al Oroantral communications closure

Introduction is within his range and which should be referred to a specialist. The flapless techniques for OACs The oroantral communication (OAC) is a patho- closure already described in the literature are found logical condition that consists of a solution of con- on positioning a material that stabilises the blood tinuity between the oral cavity and the maxillary clot into the fresh extraction socket. Some exam- sinus1,2 and usually occurs after the extraction of ples are the thermally moulded poly(lactide-co-gly- maxillary posterior teeth3. The roots of these teeth colic) acid-coated porous `-tricalcium phosphate are frequently close to the maxillary sinus, since it (RootReplica, Degradable Solutions, Schlieren, occupies a voluminous portion of the body of the Switzerland)18, the biodegradable polyurethane maxilla. In physiologic conditions, the thickness of foam20 (Polyganics BV, Groningen, The Nether- the maxillary sinus is from 1 mm to 7 mm4, but lands) and the resorbable haemostatic gauze com- the age and local pathological processes may thin posed of reconstituted oxidised cellulose19 (Surgi- the cortical bone, even causing its erosion5,6. The cel; Johnson and Johnson, Somerville, NJ, USA). incidence of OAC after extraction of premolars and The application of these materials does not require molars has been stated being between 0.3% and particular technical ability and it seems reasonable 13.0%7-11. that the majority of the general dentists can safely Spontaneous closure of the defect has been perform these flapless treatments of OACs. reported, especially when its diameter is less than 5 In order to delineate useful guidelines for the mm12, but it is still unpredictable. Simply suturing update and the clinical practice of the general the gingiva over the socket5 can treat small-sized dentist21,22, we looked for systematic review and OACs. However, if a defect remains pervious for meta-analysis concerning the treatment of OACs. days or an infection is present in the surgical site, Nowadays, only one meta-analytic study2 on the there is a higher risk of developing a chronic inflam- surgical therapy of OACs has been published. How- mation of the sinus membrane and an oroantral ever, it is peculiar that: (i) although the article was fistula may form2. In order to avoid the aforemen- published in 2011, the bibliographic research was tioned complications, a multitude of surgical tech- limited to 2008; (ii) the research strategy of the niques for immediate or delayed OAC closure has articles was not clearly specified; in fact the text did been described1. Rehrmann’s flap13 is one of the not report the algorithms that were employed, the most widespread and effective techniques, which inclusion/exclusion criteria or information about allowed more than a 90% success rate already in the review operators. As a consequence, only a the 1970s14. Nonetheless, the closure of an OAC few criteria of the AMSTAR questionnaire23 for the with flap elevation has some disadvantages: (i) it evaluation of the quality of systematic review are requires that the operator has proper experience in met by this study. oral surgery; (ii) in some cases specific instruments The aim of this systematic review was to assess and materials are necessary; (iii) the patient can be the effectiveness of different techniques for the more prone to postoperative swelling and pain; (iv) closure of OACs against nothing or other tech- sometimes a decrease of vestibular depth is una- niques, and compare the effectiveness of the tech- voidable1,15. The possibility of limiting the surgical niques in terms of rate of successful closure of the trauma and, consequently, the overall discomfort defect. Only randomised controlled studies with for the patient has raised the interest in flapless at least 2 months’ follow-up were considered. The surgical techniques16-19, which are less complex present review followed the criteria of the Preferred and more conservative, but their efficacy has still Reporting Items for Systematic Reviews and Meta- to be ascertained. This can be particularly relevant Analyses, the PRISMA statement24 (http://prisma- in case of small-sized non-complicated OACs fol- statement.org/). lowing tooth extraction, in which flap-based surgical techniques might constitute an overtreatment. No guidelines for the treatment of OACs are available to lead the dentist in the choice of which case

Eur J Oral Implantol 2014;7(4):347–357 De Biasi et al Oroantral communications closure 349

Table 1 Research algorithms used for each electronic database.

Database Web address Algorithm PubMed http://www.ncbi.nlm.nih.gov (‘Oroantral Fistula’[Mesh]) OR OAC OR OAP OR OAF OR oroantral OR oro-antral OR antrooral OR antro-oral OR orosinusal OR oro-sinusal AND (communication OR perforation OR fistula) SciVerse Scopus http://www.scopus.com TITLE-ABS-KEY(((Oroantral Fistula) OR (OAC) OR (OAP) OR (OAF) OR (oroantral) OR (oro-antral) OR (antrooral) OR (antro-oral) OR (orosinusal) OR (oro-sinusal)) AND ((communication) OR (perforation) OR (fistula))) LILACS http://lilacs.bvsalud.org/en (OAC or OAP or OAF or oroantral or oro-antral or antrooral or antro-oral or orosinusal or oro-sinusal) AND (communication or perforation or fistula) SciELO http://www.scielo.org (OAC or OAP or OAF or oroantral or oro-antral or antrooral or antro-oral or orosinusal or oro-sinusal) AND (communication or perforation or fistula) Cochrane Library http://www.thecochranelibrary.com (OAC or OAP or OAF or oroantral or oro-antral or antrooral or antro-oral or orosinusal or oro-sinusal) AND (communication or perforation or fistula)

Materials and methods illofacial Surgery; The International Journal of Oral and Maxillofacial Surgery; The Journal of Cranio- Search methods for identification of Maxillo-facial Surgery; The Journal of Oral and Max- studies illofacial Surgery; Oral Surgery Oral Medicine Oral Pathology Oral Radiology and Endodontology. Two calibrated reviewers acted independently during each phase of the review. In the case of discrepan- Study selection cies between the findings of the two reviewers, they discussed to reach a consensus. After the removal of duplicates, the two reviewers The inclusion and the exclusion criteria to con- independently and in duplicate in a non-blind fash- sider the studies for this review are randomised ion examined the title, keywords and abstract of controlled trials (RCTs) comparing patients hav- reports identified from electronic searching for evi- ing an oroantral communication treated or not dence of three criteria: with a closure technique or with different closure t Is it a randomised controlled trial? techniques with at least a follow-up of 2 months. t Does it involve the treatment of OACs? The following databases were searched for rele- t Is the success rate, the incidence of complica- vant trials: PubMed; SciVerse Scopus; Latin Ameri- tions, the patient’s discomfort and the vestibu- can and Caribbean Health Sciences (LILACS); The lar depth reduction assessed after completion of Scientific Electronic Library Online (SciELO); and treatment? The Cochrane Library. Publications from January 1949 to August 2014 were included. No language If the article met these criteria or, alternatively, if one restrictions were included in the search strategy. or both review authors were incapable of assess- Table 1 summarises the consulted databases and ing this information from the title, keywords and the research algorithms. abstract, the full text article was obtained (Fig 1). In addition, further manual searches were per- A radiographic follow-up was not considered formed considering: (i) the related citations of the strictly necessary, because the clinical examination articles selected by the PubMed database consulta- is more specific and sensible25. A minimum follow- tion; (ii) the references of the selected articles; (iii) the up period of 2 months was required for the study articles published during the past 10 years of the fol- to be included and it was established according to lowing journals: The British Journal of Oral and Max- previously published data, which demonstrate that

Eur J Oral Implantol 2014;7(4):347–357 350 De Biasi et al Oroantral communications closure

responding author within 7 days, a second email Records identiﬁed through database searching and other sources was sent and we awaited a response for a further 7 days. If the corresponding author did not answer Manuel PubMed Scorpus LILACS Scielo Cochrane the phone, a new attempt was made after 7 days. In Search (n = 972) (n = 1132) (n = 56) (n = 11) (n = 7) (n = 2) case of failure of this procedure, missing information was considered not reported. Email was chosen as

Identiﬁcation method to obtain unpublished data, since it requires the fewest number of attempts and the shortest time to respond27.

1256 records after duplicates removed Quality assessment

The following criteria were examined: Screening 1. Method of allocation concealment 1256 records screened 1251 records excluded a. Criterion ‘met’: Patients’ recruitment and assignment were randomised and the researcher recruiting participants was unaware of the randomisation process. 5 full-text articles assessed for eligibility 3 full-text articles excluded b. Criterion ‘unclear’: Such information is not

Eligibility reported. c. Criterion ‘unmet’: The randomisation schedule was not kept concealed to the researcher 2 studies included in the recruiting participants. qualitative synthesis 2. Protection against detection bias a. Criterion ‘met’: The researcher assessing the treatment outcomes was kept ‘blind’. Included 2 studies included in b. Criterion ‘unclear’: Such information is not quantitative synthesis reported. c. Criterion ‘unmet’: The researcher was not Fig 1 Flow diagram of study inclusion. ‘blind’ to the outcomes. 3. Protection against attrition bias treated OACs that are still closed after 2 months can a. Criterion ‘met’: No drop-outs or withdrawals be considered stable and healed26. took place and all outcome data are reported. The primary outcome measures were the rate Alternatively, missing outcome data are evenly of successful closure of OACs and the incidence of distributed among groups and missing for sim- complications. The secondary outcome measures ilar reasons. were the patient’s discomfort and the vestibular b. Criterion ‘unclear’: Such information is not depth reduction. reported. c. Criterion ‘unmet’: Relevant outcome data are not reported and/or missing data are imbal- Data extraction anced in either number or reasons among The two reviewers performed the data extraction groups. independently by ﬁlling out a previously prepared 4. Validity and reproducibility of the methods used form. If relevant data were missing, we attempted for the assessment of the outcomes. to contact the authors in order to obtain them. The a. Criterion ‘met’: There is a clear description of procedure began with sending an email to the corre- the methods of the primary and secondary sponding author and subsequently making a phone outcome assessment and their reproducibility call. If no email reply was obtained from the cor- has been evidenced.

Eur J Oral Implantol 2014;7(4):347–357 De Biasi et al Oroantral communications closure 351

b. Criterion ‘unclear’: Such information is not Investigation of publication bias reported. c. Criterion ‘unmet’: The outcome assessment If there were at least ten RCTs included in a meta- procedures are not described and/or no analysis of the present review, a funnel plot of effect attempt is reported for testing their reprodu- size versus standard error was to be drawn in order cibility. to assess its possible asymmetry. The analysis of such a plot allows for the evaluation of the likely presence, The validity of the studies was established by catego- or apparent absence, of publication bias and other rising each one as follows: biases. 1. Low risk of bias: All of the criteria are met. 2. Unclear risk of bias: One or more criteria unclear, the others are met. Results 3. High risk of bias: one or more criteria unmet. The research individuated 1256 studies; review of Furthermore, methodological aspects were assessed: title and abstract caused the exclusion of 1251 of t Sample size calculation reported or not. them. Full articles were retrieved for the remain- t Similar baseline characteristics of groups in order ing five, which were all written in English. Three of to allow comparability: age; gender; location and these publications did not fulfil the inclusion criteria dimension of the OAC; maxillary sinus morph- of the present review. The study by Amaratunga28 ology. was the oldest reference assessed for eligibility (pub- t Clarity of inclusion and exclusion criteria for lished in 1986) and was excluded because, even if it enrolling patients. was designed as a controlled trial, no attempt was made to allocate the patients to intervention groups in a random manner. The article by Batra et al29 was Data synthesis found not to be a RCT. Nezafati et al30 performed a For binary outcomes, the estimate of the effect of RCT comparing the effectiveness of the Rehrmann’s a surgical treatment was to be expressed as odds flap with the pedicled buccal fat pad flap; neverthe- ratios (OR) together with 95% confidence intervals less, this study was excluded because the follow- (CIs). For continuous outcomes, mean differences up was limited to 30 days postoperatively. Despite and standard deviations was to be gathered to com- extensive searching of the literature, the studies by bine and summarise the data for each group using Gacic et al19 and Hariram et al31 were the only two mean differences and 95% CIs. The patient was to that could be included in the present review. We be regarded as statistical unit. were unable to collect from the text all the informa- A Cochran’s test for heterogeneity was to be used tion needed for data extraction and quality assess- to assess the significance of any difference in the esti- ment, thus we contacted the authors by means of mates of the treatment effects from the studies. P val- both phone calls and email. ues lower than 0.1 were to be considered expression Table 2 reports the data obtained from the of significant heterogeneity, which was to be quanti- included studies regarding the primary and second- fied by means of I2 statistic, regarding I2 >50% as ary outcomes describing them in detail; items con- indication of moderate to high heterogeneity. cerning missing information that was gathered by A meta-analysis was to be performed only in the contacting the authors are marked by an asterisk. presence of studies of similar comparisons making Gacic et al19 enrolled 30 patients in a RCT to assess use of identical outcome measures. If more than the effectiveness of the closure of OACs with resorb- three studies were suitable for inclusion in the meta- able PLGA-coated `-TCP root analogues or haemo- analysis, random-effect models (program Metareg, static gauze, using the Rehrmann’s buccal flap as Stata version 9.0, Stata Corporation, USA) was to control. All 30 patients were followed up to a period be used to summarise ORs and mean differences for of 6 months without withdrawals or drop-outs. The binary and continuous data, respectively. researchers reported that all the OACs (30/30, one

Eur J Oral Implantol 2014;7(4):347–357 352 De Biasi et al Oroantral communications closure

Table 2 Characteristics of the included studies. Gacic et al19, 2009 Design Randomised controlled trial Sample size 30 patients divided into three treatment groups (n = 10) Groups Group 1 (n = 10) Group 2 (n = 10) Group 3 (n = 10) Age* 40 ± 15 y 32 ± 18 y 39 ± 16 y Technique for OAC closure `-TCP Haemostatic gauze Buccal flap Success rate 100% 100% 100% Incidence of complications 0% 0% 0% Odds ratio compared to the control 1.00 [0.02;55.27] 1.00 [0.02;55.27] / group (buccal flap) [95% CI] Z = 0.000 Z = 0.000 P = 1.000 P = 1.000 Patient discomfort assessment – 1st day 6.4 [3.3;9.5] 1st day 8.6 [6.0;11.2] 1st day 19.3 [16.9;21.7] Pain intensity 2nd day 2.5 [0.8;4.2] 2nd day 3.8 [1.9;5.7] 2nd day 11.4 [7.7;15.1] (VAS from 0 to 100) mean[95% CI] 7th day 0 [0;0] 7th day 0 [0;0] 7th day 2 .8 ± [0.9;4.7] Patient discomfort assessment – Baseline 15.1 [14.8;15.7] Baseline 14.6 [13.7;15.5] Baseline 15.3 [14.6;16.0] Extraoral swelling Postopera- 15.1 [14.8;15.7] Postoperative 14.6 [13.7;15.5] Postoperative 15.3 [14.6;16.0] (distance, in centimetres, between tive 15.1 [14.8;15.7] 1st day 14.9 [14.0;15.8] 1st day 16.5 [15.8;17.2] intertragic notch and corner of the 1st day 15.1 [14.8;15.7] 7th day 14.6 [13.7;15.5] 7th day 15.5 [14.8;16.2] mouth) mean[95% CI] 7th day Vestibular depth reduction meas- Baseline 12.7 [11.6;13.8] Baseline 13.6 [12.7;14.5] Baseline 13.4 [12.3;14.5] urement (distance, in millimetres, Post- 12.6 [11.6;13.6] Postoperative 13.6 [12.7;14.5] Postoperative 11.7 [10.6;12.8] between marginal gingiva and the operative 12.7 [11.6;13.8] 7 days 13.6 [12.7;14.5] 7 days 10.9 [9.7;12.1] highest point in the vestibule) mean 7 days 12.7 [11.6;13.8] 6 months 13.6 [12.7;14.5] 6 months 12.2 [11.0;13.4] [95% CI] 6 months Quality assessment Item Reviewers’ judgement Support for judgement Method of allocation concealment* criterion met Quote: “… 30 patients, with OACs created after tooth extraction, were randomly selected into 3 groups.” After contacting the authors, they stated that patients were randomly assigned following simple randomisation procedures (computerised random numbers) to experimental groups. The researcher recruiting participants was blind to the randomisation process. Protection against detection bias* criterion met Not mentioned. After contacting the authors, they stated that the operator assessing outcomes was blind to the randomisation schedule and did not take part to the interventions for OAC closure. Protection against attrition bias* criterion met Not mentioned. After contacting the authors, they stated that no with- drawal/drop-out was registered. Reproducibility of the methods for criterion unmet Quote: the assessment of outcome “The pain intensity was quantified using a visual analogue scale that ranged from 0 to 100 arbitrary units.” “Swelling was measured extraorally by assessing the distance between the intertragic notch and the corner of the mouth.” “The vestibular depth was measured with a compass as the distance between the marginal gingiva of the extraction site to the highest point in the vestibule in the same frontal plane.” Comment: the reproducibility of the assessment of both swelling and vestibular depth could be hindered by the resilience of soft tissues. After contacting the authors, they stated that they did not test the reproducibility of these methods. Risk of bias high Assessment of other methodological aspects Item Description Sample size calculation* After being contacted, the authors declared that sample size was not calculated Similarity of baseline conditions The authors stated that they tested the absence of differences among groups in terms of OAC diameter, among groups* gender, age, tooth originating the OAC and root originating the OAC in case of molar teeth. Clarity of inclusion and exclusion Inclusion criteria: thirty patients in good general health with OACs created after tooth extraction. criteria Exclusion criteria: smokers and pregnant or lactating woman and patients under any medication.

Eur J Oral Implantol 2014;7(4):347–357 De Biasi et al Oroantral communications closure 353

Hariram et al31, 2010 Design Randomised controlled trial Sample size 20 patients divided into two treatment groups (n = 10) Groups Group 1 (n = 10) Group 2 (n = 10) Age 37.0 ± 5.7 y Comment: information per group not available because the author did not reply Technique for OAC closure Sandwich graft with hydroxyapatite crystals Buccal fat pad flap within collagen sheaths Success rate 90% Incidence of complications 10% (1 patient) Quote: “graft was rejected by 100% first week, however no further graft rejection 0% took place” Odds ratio compared to the control 3.32 [0.12;91.6] / group (buccal fat pad flap) [95% CI] Z = 0.708 P = 0.479 Patient discomfort assessment – Baseline 7.6 [7.0;8.2] Baseline 7.3 [6.8;7.8] Pain intensity 1 week 3.9 [3.1;4.7] 1 week 3.5 [3.1;3.9] (VAS from 0 to 10) 3 weeks 2.3 [1.4;3.2] 3 weeks 1.7 [1.3;2.0] mean[95% CI] 6 weeks 1.1 [0.4;1.8] 6 weeks 1.0 [0.6;1.4] 12 weeks 0.4 [-0.1;0.9] 12 weeks 0.3 [-0.1;0.7] Patient discomfort assessment – 1 week 100% 1.00 [0.02;55.27] 1 week 100% Extraoral swelling prevalence and 3 weeks 70% 0.10 [0.00;2.28] 3 weeks 100% odds ratio compared to the control 6 weeks 20% 1.00 [0.11;8.95] 6 weeks 20% group (buccal fat pad flap) [95% CI] 12 weeks 0% 1.00 [0.02;55.27] 12 weeks 0% Vestibular depth reduction measure- Not performed ment Quality assessment Item Reviewers’ judgement Support for judgement Method of allocation criterion unclear Quote: “The patients were taken randomly irrespective of sex, caste and creed…” concealment Other information concerning the method of allocation concealment is not available because the author did not reply Protection against detection bias criterion unclear Not mentioned Information concerning the blindness of the researcher assessing the treatment outcomes is not available because the author did not reply Protection against attrition bias criterion unclear Not mentioned Information concerning withdrawals or drop-outs is not available because the author did not reply Reproducibility of the methods for criterion unclear Pain was measured by means of a visual analogue scale from 0 to 10 the assessment of outcome Vestibular depth changes were not taken into account Information concerning the methods to assess swelling is not available because the author did not reply Risk of bias unclear Assessment of other methodological aspects Item Description Sample size calculation Not mentioned Similarity of baseline conditions Not mentioned among groups Clarity of inclusion and exclusion Inclusion criteria: twenty patients with an OAC taken randomly irrespective of sex, caste and creed criteria Exclusion criteria: patients suffering from renal or hepatic disease, heart disease, blood dyscrasia, gastric ulcer, hypersensitivity, allergy or idiosyncratic reactions to any medication

Abbreviations: CI, conﬁdence interval; OAC, oroantral communication; VAS, visual analogue scale; `-TCP, thermally moulded poly(lactide-co-glycolic) acid–coated porous`-tricalcium phosphate; * , information not expressed in the text retrieved by contacting the authors.

Eur J Oral Implantol 2014;7(4):347–357 354 De Biasi et al Oroantral communications closure

per patient) were successfully closed after 6 months als that fulfilled the inclusion criteria of the present without complications, thus finding no differences in review reported optimistic results, with both flap- the success rate obtained with the three techniques less and flap approaches leading to successful OACs (100% in all groups). The patients’ discomfort was closure in almost the totality of patients19,31. The assessed in terms of pain intensity expressed as a main conclusions of Gacic et al19 are that there was visual analogue scale from 0 to 100 arbitrary units similar effectiveness between flap and flapless tech- and swelling as the distance between the intertragic niques and the latter are associated with lower pain notch and the corner of the mouth. intensity, postoperative swelling and no vestibular The trial by Hariram et al31 involved 20 patients depth reduction. Hariram et al31 found that a classic with OACs, who were randomly assigned to two technique – the buccal fat pad – and a more recent intervention groups, and compared the effectiveness technique – a sandwich graft approach – offered of a sandwich graft technique with hydroxyapatite nearly the same effectiveness, with similar patient crystals within collagen sheaths and the buccal fat discomfort. pad flap. They followed up the patients for 3 months Even if such findings appear promising, some and found all the patients to be successfully healed points of criticism concerning several methodologi- at the end of the observation period, with the only cal aspects of these works should be analysed and exception of one subject (10%) belonging to the deserve discussion. Firstly, we must underline that sandwich graft group. The difference in success rate the vast majority of the information needed to carry was reported to be not statistically significant. The out the quality assessment was not already expressed authors did not consider vestibular depth changes in the text of the articles. The corresponding au- caused by the two types of intervention. Pain was thor of the work by Hariram et al31 did not reply scored by means of a visual analogue scale from 0 to our inquiries and the risk of bias of this trial was to 10 and the presence of swelling was registered as judged as being unclear. The missing information in a binary datum at each recall. the study by Gacic et al19 was eventually retrieved Since relevant information for the quality assess- by contacting the authors. The Cochrane Handbook ment was missing in both included trials (mainly: for Systematic Reviews of Interventions advises that methods for randomisation, blindness of operators contacting authors of trial reports may lead to overly assessing outcome, occurrence of drop-outs and positive answers32; therefore, this methodological withdrawals), we could perform the quality assess- information must be handled and interpreted with ment only after having contacted the correspond- care. Furthermore, the sample size was not calcu- ing authors and retrieved the missing information. lated by the authors and this has the potential to This was feasible only for the trial by Gacic et al19, yield statistical errors. In the reviewers’ opinion, one because we were unable to get in contact with the major concern about this work is the reliability of the corresponding author of the work by Hariram et al31, methods used to assess postoperative swelling and despite multiple attempts. According to the analysis the variation of vestibular depth. After being asked, of each item reported in Table 2, the risk of bias was the authors declared that they did not test the re- judged high in the former and unclear in the latter. producibility of these methods, namely measuring the distance between the intertragic notch and the corner of the mouth with a flexible measuring tape Discussion for extraoral swelling and the vestibular depth with a drawing compass. The precision of these devices After the screening, only two19,31 of the studies provide appears questionable because of the resil- fulfilled the criteria for qualitative and quantitative ience of soft tissues, since it is likely that the applica- synthesis. Consequently, caution must be exercised tion of different pressure on both skin and the buccal when drawing a conclusion from the present review, sulcus mucosa does not allow for reliable readings. both in consideration to its primary outcome (suc- In light of these considerations, we must empha- cess rate) or its secondary outcomes (patient’s dis- sise the need for further well-conducted randomised comfort, vestibular depth reduction). The two tri- controlled trials with low risk of bias. For publications

Eur J Oral Implantol 2014;7(4):347–357 De Biasi et al Oroantral communications closure 355

included in the future, a series of tasks would be Thoma et al18. They adopted this technique that accomplished. Data to be gathered would include: consists of the chairside fabrication of root analogues technique for OAC closure; sample size and age of from freshly extracted teeth and that had previously the patients; outcome measures (success rate, inci- been described for socket preservation in implant dence of complications, patient’s discomfort, vestib- dentistry35,36. Among other materials suggested for ular depth reduction). All assessments of outcomes grafting in the socket in case of OAC, conical bio- will be recorded. The pooling of data and meta- degradable polyurethane foam has led to promising analysis will only be feasible if the selected studies results in animal models37,38 and consequently in are sufficiently similar in terms of patients’ character- humans20,39. Only one study19 assessed the success istics, OAC closure techniques and outcomes. rate of positioning resorbable haemostatic gauze in The electronic searching method made use of the socket for OAC treatment, with the intent to algorithms with nonspecific terms in order to max- stabilise the blood clot with a common and cheap imise the sensibility of the research. Since no specific material. Despite the remarkable effort to test these terminology about the techniques for the closure new conservative techniques, the study by Gacic et of OACs has been universally adopted, the biblio- al19 is the only randomised controlled trial that has graphic research was intentionally designed to be been published on this topic at the moment. sensible, even at the expense of its specificity. The The treatment conundrum concerning the small- queries involved terms related to the only concept sized OACs is relevant because they are frequent of ‘oroantral communication’ and the relative syno- and there are no guidelines for their treatment; a nyms, without referring to treatment procedures. It is study reported a 13% incidence rate of OACs after noteworthy that all the studies individuated with the maxillary third molar extraction (134/1057 extrac- manual search had already been found in the elec- tions), with the 83% of the OACs presenting a diam- tronic databases. This algorithm design individuated eter less than 3 mm (111/134 extractions)11. Other a large amount of publications (1256), but only five factors that may potentially affect the spontaneous of them19,28-31 were assessed for eligibility. healing of OACs are the sinus morphology and the A narrative review about the treatment of OACs localisation of sinus septi; a self-maintaining space is already available1, but this type of analysis is more effect might be favoured by the sinus anatomy. prone to subjective variations in research strategy33. However, to the best of our knowledge, these fac- The authors focused on the different surgical tech- tors have not been considered by researchers. Sev- niques and concluded that the treatment of election eral authors disagree on the need to treat OACs: is represented by buccal or palatal flaps. The classic some of them advocate the sole suturing of the closure techniques with flap elevation, by far the gingiva in case of OACs less than 5 mm in diam- most studied, can carry the frequent disadvantage of eter3,40, other ones regard the surgical intervention second intention healing and soft tissue topography always indicated19,41,42. It must be underlined that alteration19. The reduction of the vestibular depth the dimensions of the OAC are difficult to estimate, could jeopardise the stability of a removable prosthe- especially before the flap elevation. For this purpose, sis18 and vestibuloplasty could be indicated in a later blunt calibrated probes have been created only for period. Moreover, the preparation, the positioning experimental reasons in order to measure the OAC and the suturing of the flap may require a long time diameter18. to be learned and sometimes to be performed18. From these considerations emerge some unre- With the aim of eliminating the need of OAC solved problems: (i) instruments for the intraopera- closure by first intention, some studies evaluated the tive measurement of the OAC diameter immediately effectiveness of alternative methods, in which the after a tooth extraction are still unavailable for the defect was filled with bone substitutes with suffi- clinician, who can only draw upon a delayed radio- cient stiffness not to disperse in the sinus16-19,34. An graphic examination, such as cone beam computed innovative technique based on `-tricalcium phos- tomography; (ii) there is not a universally accepted phate root analogues (RootReplica, Degradable dimensional classification of OACs (e.g. small, Solutions, Schlieren, Switzerland) was described by medium, large) and this nosological lacuna might

Eur J Oral Implantol 2014;7(4):347–357 356 De Biasi et al Oroantral communications closure

partially justify the absence of treatment guidelines; References (iii) studies on the closure of OACs present a bias in relation to the treatment of defects that could heal 1. Visscher SH, van Minnen B, Bos RR. Closure of oroantral communications: a review of the literature. J Oral Maxillofac spontaneously, with the risk of overestimating the Surg 2010;68:1384–1391. effectiveness of the treatment. 2. Franco-Carro B, Barona-Dorado C, Martinez-Gonzalez MJ, Rubio-Alonso LJ, Martinez-Gonzalez JM. Meta-analytic The progression of the periodontal disease an in study on the frequency and treatment of oral antral commu- apical direction determines the vertical attachment nications. Med Oral Patol Oral Cir Bucal 2011;16:e682–687. 3. Abuabara A, Cortez AL, Passeri LA, de Moraes M, Moreira loss, also at the bone’s expense. This fact can gain RW. Evaluation of different treatments for oroantral/oro- further relevance in the posterior maxillary area, due nasal communications: experience of 112 cases. Int J Oral Maxillofac Surg 2006;35:155–158. to the complications that may occur. The possibility 4. Skoglund LA, Pedersen SS, Holst E. Surgical management that deep periodontal defects may elicit a sinus infec- of 85 perforations to the maxillary sinus. Int J Oral Surg tion and consequently an OAC has been described43. 1983;12:1–5. 5. Fragiskos F. Oral surgery. Berlin: Springer, 2007:190–191. Furthermore, if a periodontal patient develops an 6. Moore U. Principles of Oral and Maxillofacial Surgery, ed 6. OAC, the risk of infection of the surgical site may Chichester: Wiley-Blackwell, 2011:127–131. 7. Beckedorf H, Sonnabend E. [The incidence of maxillary be increased because of the presence of periodontal sinus perforations in teeth extractions]. Zahnarztl Rundsch pathogens, analogously to the case of dry socket 1954;63:566–569. 8. Punwutikorn J, Waikakul A, Pairuchvej V. Clinically signifi- 44 in periodontal patients . It seems appropriate to cant oroantral communications - a study of incidence and register in future clinical trials the presence of peri- site. Int J Oral Maxillofac Surg 1994;23:19–21. 9. Hirata Y, Kino K, Nagaoka S, Miyamoto R, Yoshimasu H, odontal defects of the tooth that originated the OAC Amagasa T. [A clinical investigation of oro-maxillary sinus- and indexes of the patient’s periodontal health (i.e. perforation due to tooth extraction]. Kokubyo Gakkai Zasshi 2001;68:249–253. plaque index, bleeding on probing). 10. del Rey-Santamaria M, Valmaseda Castellon E, Berini For the design of randomised controlled trials, Aytes L, Gay Escoda C. Incidence of oral sinus communications in 389 upper thirmolar extraction. Med Oral Patol Oral the problems related to the small sample size might Cir Bucal 2006;11:E334–338. be overcome by instituting multicentre research tri- 11. Rothamel D, Wahl G, d’Hoedt B, Nentwig GH, Schwarz F, als: this fact would increase the sample size, limiting Becker J. Incidence and predictive factors for perforation of the maxillary antrum in operations to remove upper wisdom the heterogeneity among the groups in terms of age, teeth: prospective multicentre study. Br J Oral Maxillofac sex, aetiology, size and localisation of the OAC, sinus Surg 2007;45:387–391. 12. von Wowern N. Correlation between the development of floor thickness, presence of periodontal defects in an oroantral fistula and the size of the corresponding bony the site of interest, and periodontal health. defect. J Oral Surg 1973;31:98–102. 13. Rehrmann R. Eine Methode zur Schliessung von Kiefer- holen Perforationen. Dtsch Zahnartl Wochenzeitschrift 1936;39:1136. 14. Killey HC, Kay LW. Observations based on the surgical clos- Conclusions ure of 362 oro-antral fistulas. Int Surg 1972;57:545–549. 15. Obradovic O, Todorovic L, Pesic V. Investigations of the buccal sulcus depth after the use of certain methods of The present systematic review assessed that at the oro-antral communication closure. Bull Group Int Rech Sci moment there are no RCTs evaluating the need of Stomatol Odontol 1981;24:209–214. OAC closure by comparing treated and untreated 16. Becker J, Kuntz A, Reichart P. Closure of oroantral perforations with hydroxylapatite ceramic. Dtsch Z Mund Kiefer patients. As to the effectiveness of different tech- Gesichtschir 1987;11:92–95. niques for OAC closure, only two trials fulfilled the 17. Zide MF, Karas ND. Hydroxylapatite block closure of oroantral fistulas: report of cases. J Oral Maxillofac Surg inclusion criteria. The two included studies reported 1992;50:71–75. that two flapless techniques and a novel sandwich 18. Thoma K, Pajarola GF, Gratz KW, Schmidlin PR. Bioabsorbable root analogue for closure of oroantral communications after graft approach can lead to success rates comparable tooth extraction: a prospective case-cohort study. Oral Surg to classic flap techniques. Also considering that the Oral Med Oral Pathol Oral Radiol Endod 2006;101:558–564. 19. Gacic B, Todorovic L, Kokovic V, Danilovic V, Stojcev-Sta- risk of bias was judged to be unclear or high, inter- jcic L, Drazic R, Markovic A. The closure of oroantral com- pretation of the findings from the present review munications with resorbable PLGA-coated beta-TCP root analogs, hemostatic gauze, or buccal flaps: a prospective must be cautious. At the moment, the choice of the study. Oral Surg Oral Med Oral Pathol Oral Radiol Endod surgical technique for the treatment of OACs cannot 2009;108:844–850. 20. Visscher SH, van Minnen B, Bos RR. Closure of oroantral be driven by a substantial amount of high-level evi- communications using biodegradable polyurethane foam: a dence and remains the prerogative of the surgeon. feasibility study. J Oral Maxillofac Surg 2010;68:281–286.

Eur J Oral Implantol 2014;7(4):347–357 De Biasi et al Oroantral communications closure

21. Oxman AD, Cook DJ, Guyatt GH. Users’ guides to the medical literature. VI. How to use an over- view. Evidence-Based Medicine Working Group. JAMA 1994;272:1367–1371. 22. Swingler GH, Volmink J, Ioannidis JP. Number of published systematic reviews and global burden of disease: database analysis. BMJ 2003;327:1083–1084. 23. Shea BJ, Grimshaw JM, Wells GA, Boers M, Andersson N, Hamel C, Porter AC, Tugwell P, Moher D, Bouter LM. Development of AMSTAR: a measurement tool to assess the methodological quality of systematic reviews. BMC Med Res Methodol 2007;7:10. 24. Moher D, Liberati A, Tetzlaff J, Altman DG, The PRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med 2009;6:e1000097. 25. von Wowern N. Clinical and radiographic fi ndings in maxillary sinus with oro-antral fi stula. Int J Oral Surg 1981;10:138–142. 26. von Wowern N. Closure of oroantral fi stula with buccal fl ap: Rehrmann versus Moczar. Int J Oral Surg 1982;11:156–165. 27. Young T, Hopewell S. Methods for obtaining unpublished data. Cochrane Database Syst Rev 2011;(11):MR000027 doi: 10.1002/14651858.MR000027.pub2 28. Amaratunga NA. Oro-antral fi stulae--a study of clinical, radiological and treatment aspects. Br J Oral Maxillofac Surg 1986;24:433–437. 29. Batra H, Jindal G, Kaur S. Evaluation of different treatment modalities for closure of oro-antral communications and formulation of a rational approach. J Maxillofac Oral Surg 2010;9:13–18. 30. Nezafati S, Vafaii A, Ghojazadeh M. Comparison of pedicled buccal fat pad fl ap with buccal fl ap for closure of oro-antral communication. Int J Oral Maxillofac Surg 2012;41:624–628. 31. Hariram, Pal US, Mohammad S, Singh RK, Singh G, Malkunje LR. Buccal fat pad versus sandwich graft for treatment of oroantral defects: A comparison. Natl J Maxillofac Surg 2010;1:6–14. 32. Higgins JPT, Green S (editors). Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 [updated March 2011]. The Cochrane Collaboration, 2011. Available from www.cochrane- handbook.org 33. Klassen TP, Jadad AR, Moher D. Guides for reading and interpreting systematic reviews: I. Getting started. Arch Pediatr Adolesc Med 1998;152:700–704. 34. Ogunsalu C. A new surgical management for oro-antral communication: the resorbable guid- ed tissue regeneration membrane--bone substitute sandwich technique. West Indian Med J 2005;54:261–263. 35. Suhonen S, Hietanen J, Marinello C, Tormala P. Custom made polyglycolic acid (PGA) - root replicas placed in extraction sockets of rabbits. Dtsch Z Mund Kiefer Gesichtschir 1995;19:253–257. 36. Suhonen J, Meyer B. Polylactic acid (PLA) root replica in ridge maintenance after loss of a vertically fractured incisor. Endo Dent Trauma 1996;12:155–160. 37. van Minnen B, Stegenga B, van Leeuwen MB, van Kooten TG, Bos RR. Nonsurgical closure of oroantral communications with a biodegradable polyurethane foam: A pilot study in rabbits. J Oral Maxillofac Surg 2007;65:218–222. 38. Visscher SH, van Minnen B, van Leeuwen MB, van Kooten TG, Bos RR. Closure of oroantral communications using biodegradable polyurethane foam: a long term study in rabbits. J Biomed Mater Res B Appl Biomater 2009;91:957–963. 39. Visscher SH, Van Minnen B, Bos RR. Feasibility of conical biodegradable polyurethane foam for closure of oroantral communications. J Oral Maxillofac Surg 2011;69:390–395. 40. Mårtensson G. Operative method in fi stulas to the maxillary sinus. Acta Otolaryngol 1957;48: 253–257. 41. Schuchardt K. Treatment of oro-antral perforation and fi stulae. Int Dent J 1955;5:157–159. 42. von Wowern N. Oroantral communications and displacements of roots into the maxillary sinus: a follow-up of 231 cases. J Oral Surg 1971;29:622–627. 43. Huang CH, Brunsvold MA. Maxillary sinusitis and periapical abscess following periodontal therapy: a case report using three-dimensional evaluation. J Periodontol 2006;77:129–134. 44. Parthasarathi K, Smith A, Chandu A. Factors affecting incidence of dry socket: a prospective community-based study. J Oral Maxillofac Surg 2011;69:1880–1884.

Eur J Oral Implantol 2014;7(4):347–357