EFFECT OF GUIDED BONE REGENERATION ON RIDGE SPLITTING WITH OR WITHOUT EXPANSION IN ADULTS: A SYSTEMATIC REVIEW

Ahmed Hanafy 1 Samar El Kholy 2 and Basma Mostafa 3 . 1. Assistant Lecturer, Department of Periodontology, Faculty of Dentistry, Beni-Suef University, Egypt. 2. Professor, Department of Periodontology, Faculty of Oral and Dental Medicine, Cairo University, Cairo, Egypt. 3. Assistant Professor, Department of Surgery and Oral Medicine, National Research Centre, Cairo, Egypt. ...................................................................................................................... Manuscript Info Abstract ......................... ........................................................................ Manuscript History

:-Studies excluded after detailed assessment of full text and the reason of exclusion.

Study
Reason of exclusion Blus and Szmukler-Moncler [10] Patients received GBR or not were included in the same group Jensen et al [32] Patients received GBR or not were included in the same group Demetriades et al [33] Patients received one-stage and two-stage procedures were included in the same group Scarano et al [27] Two-stage procedure were used Montero et al [34] Patients received GBR or not were included in the same group Anitua et al [35] Patients received GBR or not were included in the same group Shibuya et al [36] The study only measures cases with complications Critical appraisal:-Risk of bias was assessed according to study design, randomized selection, specification of the inclusion/exclusion criteria, reporting of lost follow-up and complications, objective evaluation and statistical analysis of the results.

Results:-
The selected 14 studies were divided according to the study design into 1 randomized controlled trial (RCT), 5 cohort studies and 8 case series studies. The articles collected were published in the period from 1992 to 2016. Data was collected from the selected articles in customized forms and tabulated as shown in table (2). Only one study (cohort-retrospective) 1 compared ridge splitting and expansion alone and with guided bone regeneration. And so, three studies did not apply guided bone regeneration while twelve studies applied it. Risk of bias in all selected articles was substantially high (table 3). This was due to the study designs which were mostly either cohort or case series, even the single randomized controlled trial included in the present review was of moderate risk of bias. It was not possible to perform a meta-analysis of the data because of the heterogeneity of the In all the selected papers the outcomes were objective regarding assessment of the implants success rate and implant supported prosthesis. Change in alveolar bone dimensions postoperatively was mentioned in some studies. Full thickness flap design was applied by Chiapasco et al 16

Group (2) Ridge splitting and expansion in conjunction with guided bone regeneration:-
Twelve studies applied ridge splitting and expansion with guided bone regeneration in the same surgical procedure (1 randomized controlled trial 17  . Regarding the methods used for ridge splitting, chisels and osteotomes have been used in 6 studies 1,8,17,19,20,23 , burs and discs have been used in 4 studies 22, 24-26 , electrical mallet has been used in 1 study 19 while piezo-electric device has been used in 3 studies 18,21,24 . Regarding the methods used for ridge expansion, chisels and osteotomes have been used in 8 studies 17, 19-23, 25, 26, wedges have been used in 1 study 18 while extension crest device has been used in another study 1 .
Time of implants loading was specified by Simion et al 8 17 reported that the mean marginal bone loss (in a follow up of 6 months postoperatively) of the labial plate in the control group (ridge-splitting technique using a full thickness mucoperiosteal flap) was found to be 2.29 mm (15.36%), while in the study group (ridge splitting technique using a partial thickness mucosal flap) it was 0.71 mm (5.89%). The mean palatal marginal bone loss in the control group was 2.48 mm (16.84%) and that in the study group it was 1.14 mm (8.99%). The results also showed the mean mesio-distal marginal bone loss in the control group to be 1.83 mm (12.21%), while that in the study group was 1.15 mm (8.77%).
Regarding the complications, Simion et al 8 reported a dehiscence of the membrane (non-resorbable) in one patient with two implant sites which occurred after 2 months of healing; the membrane was removed because of severe inflammation at the margins of the exposed area. Sethi and Kaus 23 reported sudden increase in the incidence of infection in 2 cases, and this practice was terminated. Kolerman et al 20 reported spontaneous exposure which occurred in 18 implants (15.5%). Spontaneous exposures were treated by replacement of the cover screw with healing abutments. In cases where there was insufficient buccal band of keratinized gingiva, masticatory mucosa pedicle flap was displaced from the palate adjacent to the implant. Membrane exposure occurred at five sites in five different patients (14%). No exposure demanded premature removal of the membrane because the exposed portions of the membrane were absorbed shortly after. As mentioned above that Tang et al 1 reported some complications but they were mentioned generally for the two study groups (ridge expansion alone and combined with guided bone regeneration), so we could not mention these complications in the results separately. The current review was conducted to focus on the effects of conjunction of guided bone regeneration with the ridge splitting with or without expansion. The results of this review showed that few studies fulfilled the inclusion and exclusion criteria with only three studies applied ridge splitting with/without expansion without the conjunction with guided tissue regeneration, while twelve studies applied guided bone regeneration, regarding that one cohort retrospective study compared both modalities and was included in both groups of the present review.
In the present review we noted heterogeneity of the identified studies regarding surgical technique (instruments and devices used in ridge splitting, flap design, flap reflection technique and bone cut extension), grafting materials and membranes as well as follow-up period. This resulted in the lack of possibility to perform a meta-analysis of the data and so the studies in this systematic review were only presented narratively. Moderate to high risk of bias was revealed for the selected studies. This is an important factor to be put into consideration when reviewing the results of this systematic review.
One study 16 was included where a device was used to expand the buccal and palatal/lingual plates after splitting; some implants were placed 7 days after the splitting procedure to allow for the activation of the expansion device within some days in the mandibular defects with very dense bone. The device used in this study could not be considered as a distraction osteogenesis device as the aim was not to generate new soft tissue and bone. Regarding the implants placed in few days after splitting, we could not refer this to staged implant placement procedure, as the aim of this step was to wait until getting enough expansion from the splitting.
In group (1), the cumulative survival rate was measured in only 1 study 1 and it was 100% with follow up period of 8 years. While in group (2), 10 studies 1,8,17,18,19,20,21,24,25,26 reported the cumulative survival rate which ranged between 97-100%. Bassetti et al 30 in their systematic review noted cumulative survival rate in 18 studies which ranged between 91.7-100%, while in another systematic review Elnayef et al 31 reported implant survival rate in between 94% to 100%. In comparison with the two groups presented in the present study, Bassetti et al 30 showed wider range of cumulative survival rate with less lower limit, while Elnayef et al 31 results were more comparable with the results of the two groups in the current review.
In group (1), the cumulative success rate of implants was measured in the three studies 1,15,16 and it ranged between 93.2-95.4% with the maximum follow up period of 8 years. In group (2), 4 studies measured the cumulative success rate 1,22,23,26 , and it was between 86.2-97% with the maximum follow up period of 8 years also. And so the results of cumulative success rates of the second group are slightly higher than the first group. Bassetti et al 30  These differences in results could be attributed to the differences in definitions of success criteria, designs of implants, inclusion and exclusion criteria, or operational details in the conducted studies.
In group (1), one study 1 reported that in the group where patients received ridge splitting and expansion alone the bone loss (vertically) after expansion was 1.61±0.91 mm after 7 months of unloaded healing, and the average of marginal bone loss after loading was 0.69±0.49 mm during the first year followed by 0.07, 0.06, 0.06, 0.06, and 0.09 mm annually. Chiapasco et al 16 reported the mean bone width variation of the expanded crest between implant placement and abutment connection, from 1-2 years after the start of prosthetic loading which were -0.4, -0.5, -0.8, and -0.8mm, respectively. While in group (2), bone loss after ridge splitting with/without expansion technique were measured in 6 studies 1,17,18,20,22,26 . Engelke et al 22 reported the mean of postoperative marginal bone loss (mesial and distal) in the follow up times of less than 1 year, 2 to 3 years and 3 to 5 years were 1.1, 2.0, and 1.9 mm respectively. Bassetti et al 18 reported the changes in peri-implant marginal bone levels between time of surgery and time of loading (4-8 months after surgery), time of loading and after 1 year, 1 year and 2 years of loading and between time of surgery and 2 years of loading and they were -1.18 ± 1.06 mm, -0.65 ± 0.98 mm, -0.05 ± 0.03 mm and -1.66 ± 1.08 mm respectively for mesial measurements, and -1.2 ± 0.99 mm, -0.53 ± 0.91 mm, -0.05 ± 0.02 mm and -1.69 ± 0.9 mm respectively for distal measurements. Kolerman et al 20 reported the follow-up time which varied between 6 and 14 months the mean vertical mesial bone loss was 1.81 mm ± 1.07, and the mean vertical distal bone loss was 1.74 mm ± 1.12. In eight patients, at least one implant presented bone loss of ≥ 3 mm. Tang et al 1 reported that in the group where patients received ridge splitting and expansion in conjunction with guided bone regeneration the bone loss (vertically) after expansion was 1.60 ± 0.81 mm after 7 months of unloaded healing, and the marginal bone loss after loading was 0.43 ± 0.51 mm during the first year, followed by 0.06, 0.08, 0.06, 0.08, and 0.09 mm annually. Garcez-Filho et al 26 documented that the marginal bone loss around implants was 0.47 ± 0.91 mm at 6 months after loading and 1.93 ± 0.93 mm at 10 years follow up after loading. Mounir et al 17 reported that the mean marginal bone loss (in a follow up of 6 months postoperatively) of the labial plate in the control group (ridge-splitting technique using a full thickness mucoperiosteal flap) was found to be 2.29 mm (15.36%) (labially), 2.48 mm (16.84%) (palatally) and 1.83 mm (12.21%) (mesio-distally). While in the study group (ridge splitting technique using a partial thickness mucosal flap) it was 0.71 mm (5.89%) (labially), 1.14 mm (8.99%) (palatally) and 1.15 mm (8.77%) (mesio-distally).
Flap design is an important factor that should be put into consideration in reviewing the results of the present contemplate. Some studies applied the partial thickness flap instead of the full thickness flap 1,17,18,19,23 . In their randomized controlled clinical trial, Mounir et al 17 reported decrease in the amount of crestal bone loss (labial, palatal and mesio-distal) in the patients where partial thickness flap were used in conjunction with split-crest procedure in comparison with full thickness flap. Tang et al 1 correlated the type of flap to be used with the severity of width insufficiency. They assumed that partial thickness flap could be used when the alveolar width is 4 mm or more while full thickness flap could be used when the width is less than 4 mm which needs guided bone regeneration in conjunction with split-crest procedure. Bassetti et al 18 stated the necessity for full thickness flap in where releasing (vertical) bone cuts are needed, as proper visualization has to be achieved. The results of the present study should be considered with caution regarding the presence of the flap design as a factor that could affect the amount of crestal bone loss postoperatively in both groups.
During reviewing complications related to the techniques used, we did not investigate fractures or cracks happened to the buccal bone plates in some cases. We think that this is an operator related not a technique related complication. Other complications happened postoperatively which was mentioned in both groups. In group (1), one case was reported with prolonged pain in the expanded area with complete resolution after 1 month postoperatively. In group (2), 2 cases were reported with sudden increase in the incidence of infection. Membrane exposure and dehiscence occurred in 6 sites in two studies 8,20 . Spontaneous exposure in 8 implants was reported in one study 20 .

Conclusion:-
In conclusion, the studies included in this review showed high success and survival rates of implants placed in narrow ridges where ridge splitting and/or expansion technique were used with/without the application of guided bone regeneration. While the conjunction of guided bone regeneration with the ridge splitting and/or expansion technique showed more complications as membranes exposure and infection. Flap design could be an important factor that could affect the alveolar bone dimensional changes postoperatively. Studies included in the present review were of high or moderate risk of bias with only one randomized controlled clinical trial. So, the results of the