Effect of Barrier Membranes on the Outcomes of Maxillary Sinus Floor Augmentation: A Meta-Analysis of Histomorphometric Outcomes Fernando Suárez-López del Amo, DDS1/Inmaculada Ortega-Oller, DDS2/Andrés Catena, PhD3/ Alberto Monje, DDS1/Vahid Khoshkam, DDS1/Laura Torrecillas-Martínez, DDS2/ Hom-Lay Wang, DDS, MS, PhD1/Pablo Galindo-Moreno, DDS, PhD2 Purpose: Sinus floor elevation via the lateral window approach represents a reliable technique for augmenting bone volume in the atrophic posterior maxilla. However, controversy remains regarding the effect of placement of a barrier membrane over the lateral window. This histomorphometric meta-analysis sought to clarify the effect of barrier membranes in lateral window sinus augmentation. Materials and Methods: An electronic search of three databases and a hand search in implant-related journals for studies published through January 2013 in the English language was conducted. Randomized controlled trials, prospective human clinical studies, retrospective investigations, and case series reporting histomorphometric results after sinus elevation using the lateral window approach with at least six patients and a minimum follow-up period of 6 months were included. Results: The initial search yielded 1,040 articles, of which 94 were further evaluated for eligibility. Finally, 37 studies were chosen and separated into membrane (group 1) and no-membrane (group 2) groups. Similar vital bone formation was found in both groups: 32.36% for group 1 and 33.07% for group 2. Conclusion: Based upon this meta-analysis, the presence of a barrier membrane over the window does not influence the amount of vital bone formation after sinus augmentation. Additionally, the type of grafting material used and healing time did not influence the histomorphometric outcome. Int J Oral Maxillofac Implants 2015;30:607–618. doi: 10.11607/jomi.3886 Key words: dental implants, dental prosthesis, endosseous implants, implant-supported prosthesis, sinus floor augmentation

A

deficiency of bone height in the posterior maxilla often poses a challenge for ideal placement of standard implants. To overcome this problem, several treatment options, such as a shortened maxillary arch,1 the use of short implants,2–6 a transalveolar technique using osteotomes,7–9 and sinus floor augmentation via a lateral approach,10,11 have been proposed.

1Graduate

Periodontics, Department of Periodontics and Oral Medicine, University of Michigan School of Dentistry, Ann Arbor, Michigan, USA. 2Department of Oral Surgery and Implant Dentistry, University of Granada, Granada, Spain. 3Department of Experimental Psychology, University of Granada, Granada, Spain. Correspondence to: Dr Fernando Suárez López del Amo, Department of Periodontics and Oral Medicine, University of Michigan School of Dentistry, 1011 North University Avenue, Ann Arbor, MI 48109-1078, USA. Email: [email protected] ©2015 by Quintessence Publishing Co Inc.

The predictability of the lateral window sinus augmentation procedure has been widely investigated; the technique has demonstrated high success rates, not only for vital bone (VB) formation, but also for the implants placed in this newly regenerated bone.12–15 Autogenous bone as well as bone substitutes, with or without barrier membranes, are commonly used for the augmentation of the maxillary sinus.16 A systematic review revealed that the use of 100% autografts or a combination of autogenous bone with other bone substitutes did not affect the survival of the implants; in addition, it was concluded that the use of barrier membranes yielded higher implant survival rates.14 The characteristics and properties of the newly formed bone have been evaluated in multiple studies. Different values have been reported for histologic and histomorphometric outcomes.13,17–21 Currently, great variability exists among the different types of grafting materials being used in sinus augmentation; however, there is no evidence to support the superiority of any particular material in terms of implant survival or absence of complications.22,23 It is The International Journal of Oral & Maxillofacial Implants 607

© 2015 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY. NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.

Suárez-López del Amo et al

generally accepted now that the type of grafting material does not have a critical effect on the outcomes of maxillary sinus augmentation. Some histomorphometric studies have shown that the application of barrier membranes inhibits soft tissue invagination, thus promoting more VB formation.24,25 Tarnow et al24 reported that the placement of an expanded polytetrafluoroethylene (e-PTFE) membrane over the lateral window defect resulted in 25.5% vital bone formation, versus 12% when no membrane was placed over the window. In addition, e-PTFE membrane coverage over the lateral window defect was associated with a 100% implant survival rate, versus 93% in sites without membrane coverage. Furthermore, Tawil and Mawla26 showed that covering the window with a collagen membrane resulted in a much higher success rate versus no membrane coverage (93.1% vs 78.1%, respectively). However, there is limited evidence in the literature to assess the effectiveness of barrier membrane placement. Hence, this histomorphometric meta-analysis on histomorphometric outcomes sought to evaluate the effectiveness of barrier membranes during lateral window sinus augmentation procedures.

MATERIALS AND METHODS Information Sources and Development of Focused Question

A search of three electronic databases—PubMed, Cochrane Central, and Ovid (MEDLINE)—for studies published until January 2013 in the English language was conducted by two examiners (FS and IO-O). The following PICO (patient, intervention, comparison, outcome) question was developed to be answered: in patients with an atrophic posterior maxilla undergoing sinus augmentation via the lateral window approach, does the presence of a barrier membrane over the lateral window affect the histomorphometric outcomes and thereby the amount of vital bone? The authors used the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-analysis)27 guidelines to determine the influence of several parameters on bone formation in the maxillary sinus. Such parameters included the use of barrier membranes as well as the grafting material employed. The PRISMA methodology was developed to ensure a more consistent study outcome. Thus, the reader can be assured that appropriate due diligence was performed during the literature search and that the search was done in a logical manner.28

Article Screening Process

The search strategy employed was a combination of free text words and Medical Subject Headings (MeSH

terms). The search terms were as follows: “sinus floor augmentation” [MeSH] OR (sinus augmentation) OR (sinus floor elevation) OR (maxillary sinus lift) OR (sinus graft) AND (membrane) OR (lateral window augmentation) OR “dental prosthesis, implant-supported” [MeSH] OR “anterior implant rehabilitations” [title or abstract (tiab)] OR “dental implantation, endosseous” [MeSH] AND (“posterior maxilla” [tiab] OR “biopsy” [tiab] OR “histology” [tiab] OR “histomorphometry” [tiab]). Additionally, a hand search of relevant studies published in dental journals between January 2000 and December 2012 was performed. The dental journals included were Journal of Clinical Periodontology, Clinical Oral Implants Research, The International Journal of Oral & Maxillofacial Implants, Implant Dentistry, Journal of Oral Implantology, Journal of Oral and Maxillofacial Surgery, Journal of Dental Research, Journal of Prosthetic Dentistry, Journal of Periodontology, The International Journal of Periodontics & Restorative Dentistry, International Journal of Oral and Maxillofacial Surgery, Clinical Implant Dentistry and Related Research, and European Journal of Oral Implantology. In addition, the reference lists of the selected studies were systematically screened for potential articles.

Eligibility Criteria

Articles that met the following criteria were included in this study: reported histomorphometric analysis, prospective human clinical trials, retrospective or case series, more than six patients in the study, and a minimum period of 6 months of healing after the lateral sinus window approach. Articles were excluded if they had one or more of the following characteristics: case reports, review articles or clinical trials with fewer than six patients, insufficient healing time between grafting and biopsy harvesting, or a lack of histomorphometric analysis. Animal studies were also excluded. Nonetheless, the bibliographies of the excluded studies were screened for potentially useful articles. Potential articles were independently reviewed in full by the two examiners. Interreviewer agreement was calculated by means of kappa statistics. Any discrepancies were resolved by discussion. The final decision on the included articles was made by mutual agreement between the two examiners. Quality assessment and data extraction of the included publications were performed independently by the two reviewers. As before, any disagreement was resolved by discussion.

Quality of the Included Studies

All studies included in the present meta-analysis, except for randomized controlled trials (RCTs), were evaluated by the Newcastle-Ottawa Scale to assess the quality of such studies to ensure a proper understanding of nonrandomized studies.29 One blinded examiner evaluated the nonrandomized trials with this method.

608 Volume 30, Number 3, 2015 © 2015 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY. NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.

Suárez-López del Amo et al

Statistical Analysis

Publication bias was assessed qualitatively and quantitatively. Qualitative analysis assessed the criteria for determining the quality of the research. The authors used the Egger test for quantitative analyses. Potential effects of the variables of interest (membrane use, type of graft material, and healing time) on the outcome variables (bone, tissue, and graft) were evaluated using an analysis of variance approach to meta-analysis.30 The random-effects model with separate Τ2 estimates was used when heterogeneity among studies was significant. Heterogeneity was computed by the Q statistic. A P value of .05 was set as the threshold for significance.

PubMed, Ovid(MEDLINE), and Cochrane databases searched Limits: English-language articles only

1,040 potential articles identified through database searching 899 articles excluded based on the titles 141 potentially relevant articles considered for further evaluation

RESULTS The screening process is summarized in Fig 1. The initial screening yielded a total of 1,040 articles. Independent screening of titles resulted in a total of 141 articles for further consideration. After examination of the abstracts, 94 articles were further evaluated in full; of these, 37 articles were selected for this metaanalysis. The kappa value for interexaminer agreement in selecting the included studies was 0.85. All the articles included in the present systematic review were controlled human clinical trials (eg, cohorts and case series) aiming to evaluate histomorphometrically/ histologically the effect of several materials for sinus elevation, except for four studies,31–34 in which patients were randomly selected. The mean Newcastle-Ottawa Scale score of the nonrandomized included studies was 6.76 ± 2.05, suggesting that some included studies could be characterized as “non-acceptable” in quality. The reasons for exclusion after full-text evaluation were: no bone graft,35 not enough patients,36–38 insufficient follow-up,17,39 case report,40 review or metaanalysis,41,42 no biopsy or histomorphometry,26,43–70 use of biologic agents that could influence bone formation,71,72 incomplete data,24,73–84 repeated patient pool,85–87 and the use of another type of graft or different technique for sinus augmentation.88,89 Therefore, from the electronic search and screening of the reference lists, 37 articles were selected.18–21,25,31–34,90–117 The manual search resulted in no additional publications. Thus, a total of 37 studies were included in this analysis. Because only four RCTs were identified, subordinate levels of evidence were included: longitudinal studies, case series, and retrospective studies. Table 1 summarizes the selected studies from 1999 to January 2013 with an observation period of 6 months to 3 years. A total of 28 studies used xenograft,19–21,25,31–34,90–109 12 used autograft,19,20,25,91, 93,99,100,105,106,109,110 4 used allograft,18,32,111,112 and 13 used alloplast19,31,33,34,90,101,102,105,110,113–117 as the sole

47 articles excluded based on the abstract evaluation 94 articles included in this review 57 articles excluded based on the full-text evaluation (see text) 37 articles included in this review Fig 1   Flow chart of the article screening process.

grafting material or in combination with other materials to fill the space created under the sinus membrane. With regard to the presence of a membrane over the lateral window, 25 articles used absorbable membranes, 18–21,25,31–34,90–95,97,99,100,103,104,108–110,112,117 3 articles used nonresorbable membranes,25,113,116 and 14 articles did not use membrane25,91,92,96,98,100–102,105–107,111,114,115 for at least one of the groups included in the study.

Publication Bias

The Egger test indicated that there was no evidence of bias effects on VB (P = .34) or remaining graft (RG) (P = .78), but there was evidence of bias effects on soft tissue (ST) (P < .01).

Heterogeneity of Studies

There was evidence to favor a relative amount of between-study heterogeneity for VB (P < .01), ST (P < .01), and RG (P < .01). Thus, a random-effect model with separate Τ2 was used to estimate the differences between the conditions of interest. The International Journal of Oral & Maxillofacial Implants 609

© 2015 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY. NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.

Suárez-López del Amo et al

Table 1  Data in the Included Studies Study Artzi et al (2001)90

Type of study PR

No. of patients 10

Artzi et al (2005)19

N/A

12

Avila-Ortiz et al (2012)18 Barone et al (2012)21 Barone et al (2013)91

CS PR PR

Boeck-Neto et al (2002)110

N/A

Cammack et al (2005)113

RT

Canullo and Dellavia (2009)114 Choi et al(2009)92

PR PR

21 24 9 9 5 5 31 13 8 6

Cordaro et al (2008)31

RCT

37

De Vicente et al (2010)93 Degidi et al (2006)94 Esfahanizadeh et al (2012)95 Felice et al (2009)96 Ferreira et al (2009)97 Froum et al (2006)32

PR N/A CP CP N/A RCT

34 10 8 10 314 11

Froum et al (2008)33

RCT

12

Fugazzotto (2003)98 Galindo-Moreno et al (2010)99 Galindo-Moreno et al (2012)20 Hallman et al (2002)100

N/A PR PR PR

N/A 45 52 11

Hanisch et al (1999)101 Iezzi et al (2011)102

PR PR

10 5 15

Lee et al (2006)103 Lee et al (2012)104 Lindgren et al (2009)34

N/A N/A RCT

10 25 11

Mangano et al (2006)115 Noumbissi et al (2005)111 Schopper et al (2003)116 Schulze-Späte et al (2012)117 Simunek et al (2008)105

N/A PR N/A CS PR

Soardi et al (2011)112 Trisi et al (2003)106 Valentini et al (2000)107 Wallace et al (2005)25

Yildirim et al (2000)108 Yildirim et al (2001)109

No. of. sinuses 10 10 12 12 21 24 9 9 5 5 31 13 8 6 6 23 25 42 10 9 10 406 11 11 12 12 N/A 90 52 11 11 14 N/A 30

No. of specimens 10 10 12 12 20 24 9 9 5 5 31 13 8 N/A N/A N/A 14 10 N/A N/A 7 11 11 21

Healing time (mo) 12 12 6 6 6 10 10 6 to 36 6 to 36 6 8 6 to 8 6 to 8 9 6 10 6 10 to 12 6.5 to 8 6 to 8

6 10 7 9.5 to 10 9

N/A N/A N/A N/A

24 6 26 6 10 10 10 10 8 23 6 15 51

11 N/A 11 11 29 6 52 7 10 10 10 10 8 28 N/A 20 64

5 N/A N/A 8 8 9 5 12 12 12 12 12 14 25 11 11 11 N/A 69 N/A N/A

12 to 13 6 6 12 to 15

12 N/A N/A N/A

9 17 12 6 to 10

N/A N/A

11 12

15 13

22 23

12 6

12 9 8

6.8 7.1

AU = autogenous; ALG = allograft; ALP = alloplast; CP = clinical pilot; CS = case series; NRS = nonresorbable; PR = prospective; RCT = randomized controlled trial; RS = resorbable; RT = retrospective; XG = xenogeneic; +/– = with or without.

610 Volume 30, Number 3, 2015 © 2015 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY. NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.

Suárez-López del Amo et al

Histomorphometric findings Graft material XG ALP XG + AU ALP+AU ALG XG AU+XG AU+XG ALP+AU AU+ALP ALP ALP ALP XG XG XG ALP XG + AU XG XG XG XG ALG XG ALP XG XG XG + AU XG + AU AU XG XG+AU ALP+XG ALP ALP ALP XG XG XG XG ALP XG ALP ALG ALP ALP ALP ALP+AU XG XG + AU AU ALG XG + AU XG XG ± AU XG ± AU XG ± AU XG XG + AU

Membrane RS RS RS RS RS RS RS None RS RS NRS NRS None RS None RS RS RS RS RS None RS RS RS RS RS None RS RS None None RS None None None None None None RS RS RS RS None None NRS RS None None None None None RS None None None RS NRS RS RS

New vital bone (%) ± SD 42.1 ± 10.01 32.2 ± 8.15 45.6 ± 10.9 32.0 ± 8.4 20.47 ± 18.25 43.95 ±18.6 30.7 ± 15.5 28.1 ± 19.4 50.46 ± 16.29 46.79 ± 8.56 41.07 ± 19.17 36.00 ± 19.12 48 ± 4.63 13 12 19.8 ± 7.9 21.6 ± 10.0 29 ± 6.6 35 30 ± 0.6 36.1 ± 4.6 38.95 ± 11.88 28.25 12.44 28.4 ± 23.8 22.3 ± 6.4 68.8 46.1 ± 16.6 35.75 ± 16.42 37.7 ± 31.3 39.9 ± 8 41.7 ± 26.6 20.7 ± 8.3 33.2 ± 1.2 30.5 ± 3.4 28.1 ± 3.9 31.8 ± 2.9 32.9 ± 0.5 26.6 ± 6.5 21 ± 6 41.1 ± 9.8 41.6 ± 14.0 38.5 ± 4.5 40.33 23.0 ± 8.3 52 ± 7.8 21.4 ± 8.1 24.0 ± 6.6 34.2 ± 13.1 24.4 ± 9.1 49.2 ± 3.1 34.75 ± 6.63 34.75 ± 5.28 27.55 ± 4.88 12.1 17.6 16.9 14.7 ± 5 18.9 ± 6.4

Nonmineralized tissue (%) ± SD 33.3 ± 14.66 43.2 ± 6.38 N/A N/A 50.47 ± 12.76 41.8 ± 22.7 50.6 ± 18.7 59.3 ± 15.4 41.31 ± 17.81 46.35 ± 12.48 51.18 ± 22.71 53.89 ± 16.97 24 ± 7.23 36 55 42.5 ± 6.9 51.8 ± 8.6 50 ± 4.5 N/A 83.8 ± 12.9 31.3 ± 5.6 52.85 ± 9.30 64.10 54.56 43.3 ± 13.2 51.7 ± 9.1 31 42.3 ± 15.1 40.56 ± 16.23 N/A N/A N/A 55.5 ± 11.1 39.3 ± 3.4 43.6 ± 2.5 45.6 ± 4.5 38.7 ± 2.7 36.4 ± 2.3 44.7 ± 7.3 N/A N/A N/A 44.6 ± 4.2 N/A 44.0 ± 13.7 N/A 39.6 ± 4.8 42.2 ± 4.0 35.0 ± 10.2 42.3 ± 8.3 50.5 ± 3.5 N/A 54.22 ± 9.83 45.44 ± 9.12 63.3 56 51.2 55.6 ± 7.0 51.5 ± 9.4

Graft (%) ± SD 24.7 ± 9.99 24.6 ± 8.37 26.0 ± 7.5 25.3 ± 5.6 29.04 ± 24.94 14.2 ± 13.6 18.4 ± 20.3 12.6 ± 12.4 8.24 ± 3.6 6.85 ± 5.55 9.56 ± 17.86 10.11 ± 12.80 28 ± 5.33 50 33 37.7 ± 8.5 26.6 ± 5.2 21 ± 7 30 16.1 ± 12.9 33.4 ± 5.6 8 ± 2.72 7.65 33.00 28.4 ± 14.9 26.0 ± 9.7 0.14 37 ± 25.1 23.69 ± 18.23 No graft remaining 12.3 ± 8.5 11.8 ± 3.6 20.9 ± 9.1 30.1 ± 0.9 28.1 ± 0.9 27.1 ± 1 33.1 ± 1.9 32.8 ± 2.1 28.7 ± 5.4 40 ± 7 10.80 12 12 ± 2.3 4.67 33.0 ± 7.8 16.1 ± 9.4 39.0 ± 7.2 33.8 ± 6.2 30.8 ± 12.4 33.3 ± 4.1 0.3 ± 0.5 5.85 ± 3.11 11.04 ± 6.03 27.01 ± 11.64 24.3 26.4 31.9 29.7 ± 7.8 29.6 ± 8.9

The International Journal of Oral & Maxillofacial Implants 611 © 2015 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY. NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.

Suárez-López del Amo et al

Study

% weight

Membrane Artzi et al (2001)90 Artzi et al (2001)90 Artzi et al (2005)19 Artzi et al (2005)19 Avila-Ortiz et al (2012)18 Barone et al (2012)21 Barone et al (2013)91 Boeck-Neto et al (2002)110 Cammack et al (2005)113 Cordaro et al (2008)31 Cordaro et al (2008)31 De Vicente et al (2010)93 Esfahanizadeh et al (2012)95 Ferreira et al (2009)97 Froum et al (2008)33 Froum et al (2008)33 Galindo-Moreno et al (2010)99 Galindo-Moreno et al (2012)20 Hallman et al (2002)100 Lee et al (2006)103 Lee et al (2012)104 Lindgren et al (2009)34 Lindgren et al (2009)34 Schopper et al (2003)116 Schulze-Späte et al (2012)117 Soardi et al (2011)112 Yildirim et al (2000)108 Yildirim et al (2001)109

2.30 2.38 2.30 2.40 2.16 2.20 1.97 2.11 2.45 2.48 2.45 2.47 2.45 2.09 1.57 2.44 1.96 1.97 1.37 2.47 2.51 2.33 2.13 2.53 2.33 2.45 2.52 2.50 63.28

Subtotal (I2 = 95.0%, P = .000) No membrane Barone et al (2013)91 Canullo and Dellavia (2009)114 Felice et al (2009)96 Hallman et al (2002)100 Hallman et al (2002)100 Hanisch et al (1999)101 Iezzi et al (2011)102 Iezzi et al (2011)102 Mangano et al (2006)115 Simunek et al (2008)105 Simunek et al (2008)105 Simunek et al (2008)105 Simunek et al (2008)105 Simunek et al (2008)105 Trisi et al (2003)106 Valentini et al (2000)107

1.75 2.48 2.50 1.08 2.35 2.22 2.52 2.54 2.50 2.38 2.44 2.15 2.34 2.53 2.43 2.52 36.72

Subtotal (I2 = 96.7%, P = .000) Overall (I2 = 96.6%, P = .000)

Membrane Artzi et al (2001)90 Artzi et al (2001)90 Avila-Ortiz et al (2012)18 Barone et al (2012)21 Barone et al (2013)91 Boeck-Neto et al (2002)110 Cammack et al (2005)113 Cordaro et al (2008)31 Cordaro et al (2008)31 De Vicente et al (2010)93 Esfahanizadeh et al (2012)95 Ferreira et al (2009)97 Froum et al (2008)33 Froum et al (2008)33 Galindo-Moreno et al (2010)99 Galindo-Moreno et al (2012)20 Lee et al (2006)103 Schopper et al (2003)116 Yildirim et al (2000)108 Yildirim et al (2001)109 Subtotal (I2 = 95.0%, P = .000) No membrane Barone et al (2013)91 Canullo and Dellavia (2009)114 Felice et al (2009)96 Hanisch et al (1999)101 Iezzi et al (2011)102 Iezzi et al (2011)102 Mangano et al (2006)115 Simunek et al (2008)105 Simunek et al (2008)105 Simunek et al (2008)105 Simunek et al (2008)105 Simunek et al (2008)105 Trisi et al (2003)106 Valentini et al (2000)107 Subtotal (I2 = 94.1%, P = .000) Overall (I2 = 93.3%, P = .000)

% weight

2.45 3.23 3.01 2.45 1.98 2.28 2.43 3.36 3.30 3.40 2.56 2.81 2.60 3.00 2.41 2.30 3.25 3.32 3.35 3.25 56.74

2.30 3.09 3.29 2.35 3.39 3.46 3.39 3.34 3.39 2.90 3.08 3.41 2.65 3.23 43.26 100.00

–92.2 92.2 NOTE: Weights are from random-effects analysis Fig 3   Effect of membrane use on ST.

100.00

–59.4 59.4 NOTE: Weights are from random-effects analysis Fig 2   Effect of membrane use on VB formation.

Study

Effects of Membrane Use

Figures 2 to 4 display forest plots summarizing the estimates of the effects of membrane vs no membrane on VB, ST, and RG under the random-effects model.

612 Volume 30, Number 3, 2015 © 2015 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY. NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.

Suárez-López del Amo et al

The results indicated that VB was similar for the nomembrane condition (33.07%) and the membrane condition (32.36%) (z = 0.25, P > .1) (Fig 2). In stark contrast, the membrane condition (48.23%) resulted in more ST than the no-membrane one (42.52%) (z = 2.39, P ≤ .02). As mentioned earlier, results for the RG variable were suspect because of extreme heterogeneity among the studies in the no-membrane condition. When separate Τ2 estimates were considered, none of the differences were significant between the two membrane conditions (22.12 vs 23.72, z = 0.54, P > .1). No attempt was made to compute differences between the two types of membranes, as only three studies25,113,116 used nonresorbable membranes.

Effects of the Type of Grafted Material

When the analysis was restricted to xenogeneic graft material (15 studies, 10 with membrane), no significant differences between membrane (33.48%) and nomembrane (29.34%) approaches were found (z = 1.25, P > .1). However, the no-membrane condition (50.71%) outperformed the membrane one (37.38%) for ST (12 studies, 7 with membrane, z = 2.93, P ≤ .01). No differences on the basis of membrane use were observed for RG (14 studies, 9 with membrane). In addition, no differences on the basis of membrane use were observed when blends of xenograft + autogenous or alloplastic material were used as the graft materials (all P > .2). There was no effect of the grafted material on VB (xenogeneic 30.84%, xenogeneic + autologous 32.95%, alloplastic 34.22%; Q(2) = 0.89, P > .6) or RG (xenogeneic 25.95%, xenogeneic + autologous 25.33%, alloplastic 24.31%; Q(2) = .14, P > .7). However, significant differences were observed among graft materials for ST (xenogeneic 46.06%, xenogeneic + autologous 47.72%, alloplastic 41.98%; Q(2) = 4.18, P = .04). Differences were observed between xenogeneic + autologous and alloplastic (P = .02) but not between xenogeneic and alloplastic (P = .12).

Effects of Healing Time on VB, RG, and ST

Four categories for healing time were used for analyses: up to 6 months, 6 to 9 months, 9 to 12 months, and more than 12 months. Association between healing time and VB was noted. VB was greater at 6 months (35.87%), 12 months (35.43%), and > 12 months (37.57%) than at 9 months (27.46%) (all P < .05). Regarding ST, healing times longer than 6 months (46.38%, 49.76%, and 49.83%) outperformed the healing time of 6 months (41.34%; all P < .03). No evidence was found favoring an effect of healing time on RG (estimates: 24.33%, 26.38%, 21.81%; Q(2) = 1.11, P > .50). In the last two analyses, the > 12 months condition was excluded because of the scarce number of studies in this category.

Study

% weight

Membrane Artzi et al (2001)90 Artzi et al (2001)90 Artzi et al (2005)19 Artzi et al (2005)19 Avila-Ortiz et al (2012)18 Barone et al (2012)21 Barone et al (2013)91 Boeck-Neto et al (2002)110 Cammack et al (2005)113 Cordaro et al (2008)31 Cordaro et al (2008)31 De Vicente et al (2010)93 Esfahanizadeh et al (2012)95 Ferreira et al (2009)97 Froum et al (2008)33 Froum et al (2008)33 Galindo-Moreno et al (2010)99 Galindo-Moreno et al (2012)20 Hallman et al (2002)100 Lee et al (2006)103 Lee et al (2012)104 Schopper et al (2003)116 Schulze-Späte et al (2012)117 Soardi et al (2011)112 Yildirim et al (2000)108 Yildirim et al (2001)109

2.43 2.45 2.47 2.49 2.27 2.45 2.17 2.49 2.51 2.49 2.50 2.48 2.36 2.50 2.33 2.43 2.06 2.25 2.50 2.50 2.50 2.51 2.41 2.51 2.49 2.48 63.04

Subtotal (I2 = 95.0%, P = .000) No membrane Barone et al (2013)91 Canullo and Dellavia (2009)114 Felice et al (2009)96 Hallman et al (2002)100 Hanisch et al (1999)101 Iezzi et al (2011)102 Iezzi et al (2011)102 Mangano et al (2006)115 Simunek et al (2008)105 Simunek et al (2008)105 Simunek et al (2008)105 Simunek et al (2008)105 Simunek et al (2008)105 Trisi et al (2003)106 Valentini et al (2000)107

2.37 2.48 2.49 2.44 2.38 2.51 2.51 2.51 2.47 2.48 2.39 2.50 2.51 2.46 2.46 36.96

Subtotal (I2 = 99.8%, P = .000) Overall (I2 = 99.6%, P = .000)

100.00

–52.6 52.6 NOTE: Weights are from random-effects analysis Fig 4   Effect of membrane use on RG.

The International Journal of Oral & Maxillofacial Implants 613 © 2015 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY. NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.

Suárez-López del Amo et al

DISCUSSION This meta-analysis summarized the currently available evidence regarding the effects of barrier membrane use in lateral window sinus floor elevation procedures on VB formation, as measured histomorphometrically. The results showed that the use of barrier membranes does not influence the amount of VB harvested at least 6 months postaugmentation; average VB was 33.07% for the no-membrane and 32.36% for the membrane group (z = 0.25, P > .1). These data disagree with the results of a previous study24 that reported a significant increase in the volume of VB (25.5%) in sites with membranes compared to no-membrane sites (11.9%) and confirms the results of some available controlled trials91,92,100 that showed that the use of a membrane did not substantially increase the amount of VB. This may be explained by the many related variables that were not properly controlled in the included studies. These variables included but were not limited to the size of the window,17 the buccopalatal dimensions of the sinus cavity,85 remaining alveolar ridge height,18 and location from which biopsy specimens were taken. For example, the location and depth of the harvested tissue can profoundly influence the mineralization and degree of bone formation observed,85 since the membrane is placed over the buccal surface of the maxilla, whereas the bone core biopsy samples were often taken from the crestal aspect of the ridge. Hence, outcome may not represent the true effect of a barrier membrane. It is also obvious that the amount of VB from a sample with 8 mm initial bone height would be different from one obtained in an area with 3 mm vertical bone height, since the sample would include both newly formed bone and transplanted residual bone. Not surprisingly, it was shown in a preclinical study that the greatest amount of bone formation takes place in the lateral areas of the sinus, while the least bone formation was noted close to the sinus membrane.118 Hence, another determining factor in the calculation of VB might be the horizontal dimension of the lateral window and the proximity of the lateral walls of the sinus to core harvesting sites. Of interest, this location was not indicated in any of the included studies. It is also important to note that some confounding factors, such as postsurgical infection, sinus membrane perforation, and patients’ systemic conditions and habits (eg, smoking), were not properly mentioned/controlled in all the studies included. All these factors might affect VB formation; however, accurate information about the incidence of complications or the population of smokers was not included in most of the studies. Hence, it is important for future studies to properly control these influential factors so that the effect of a barrier membrane in the lateral sinus window approach can be properly determined.

To date, sinus augmentation remains the most predictable procedure for the treatment of the atrophic posterior maxilla for implant-supported rehabilitations.14,15 Autologous bone graft has long been considered the gold standard of grafting materials; however, in the past decade, several systematic reviews48,119 have shown that bone replacement grafts represent a reliable alternative in terms of implant survival rates.22 It should be noted that evidence-based reviews used data from multiple and often diverse studies. Hence, it is difficult, if not impossible, to establish a large database without considering confounding variables. For instance, when evaluating the efficacy of bone substitutes, the researcher cannot ignore the effect of implant microstructure. In fact, in a recent systematic review,15 after machined-surface implants were excluded from analyses, the outcomes with different grafting materials were revealed to be similar. The results of the aforementioned reviews substantiate the finding that implant survival with bone substitutes is just as good if not better than that achieved with autologous bone.14 This meta-analysis supports the aforementioned findings, since no statistically significant differences were found between different grafting materials in terms of the amount of VB in grafted sinuses. These results are also in agreement with previous studies showing successful outcomes with a great variety of biomaterials.120,121 Although there was no effect of the type of grafting materials upon VB formation, a difference was noted in the percentage of connective tissue/bone marrow among the different types of graft materials used. Differences were observed between xenogeneic + autologous and alloplastic grafts, but not between xenogeneic and alloplastic grafts. When the healing time was taken into account, the results showed that a longer healing time resulted in more VB formation. When different time intervals (up to 6 months, 6 to 9 months, 9 to 12 months, and more than 12 months) were considered, the most active period of bone formation was found to be between the time of surgery and the following 6 months. In addition, 9-month biopsy samples showed less VB formation in comparison to the other healing periods. This was in agreement with a previous review that found more total bone volume at less than 4.5 months in comparison to between 4.5 and 9 months for autologous bone particles.122 Limitations of the study include the analysis of case series and retrospective studies, along with prospective studies and RCTs, with different sample sizes and short follow-up periods. Also, several variables that could influence the results were not reported. More controlled studies are warranted to explore the potential advantages and disadvantages of the use of

614 Volume 30, Number 3, 2015 © 2015 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY. NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.

Suárez-López del Amo et al

membranes in sinus floor augmentation procedures to properly evaluate the quality and quantity of reconstructed bone.

CONCLUSIONS Within the limitations of this meta-analysis, it can be concluded that vital bone formation after sinus augmentation using the lateral window approach is not influenced by the use of barrier membranes over the window, the grafting material, or the healing time. Therefore, it is assumed that the most relevant factor for the maturation of the graft material is the recipient site, whereas the choices of biomaterial and membrane remain secondary factors.

ACKNOWLEDGMENTS This study was partially supported by the University of Michigan Periodontal Graduate Student Research Fund. The authors report no conflicts of interest related to this study.

REFERENCES 1. Fontijn-Tekamp FA, Slagter AP, Van Der Bilt A, et al. Biting and chewing in overdentures, full dentures, and natural dentitions. J Dent Res 2000;79:1519–1524. 2. Van Assche N, Michels S, Quirynen M, Naert I. Extra short dental implants supporting an overdenture in the edentulous maxilla: A proof of concept. Clin Oral Implants Res 2012;23:567–576. 3. Pommer B, Frantal S, Willer J, Posch M, Watzek G, Tepper G. Impact of dental implant length on early failure rates: A meta-analysis of observational studies. J Clin Periodontol 2011;38:856–863. 4. Sanchez-Garces MA, Costa-Berenguer X, Gay-Escoda C. Short implants: A descriptive study of 273 implants. Clin Implant Dent Relat Res 2012;14:508–516. 5. Monje A, Chan HL, Fu JH, Suarez F, Galindo-Moreno P, Wang HL. Are short dental implants (

Effect of barrier membranes on the outcomes of maxillary sinus floor augmentation: a meta-analysis of histomorphometric outcomes.

Sinus floor elevation via the lateral window approach represents a reliable technique for augmenting bone volume in the atrophic posterior maxilla. Ho...
268KB Sizes 0 Downloads 18 Views