Experimental design and treatment protocols
In this prospective, split-mouth, randomized controlled clinical study, 28 subjects (21 - 40 years old) requiring correction of EGD will received ECL using OF and FL techniques. A computer-generated table randomly distributed the right quadrant to receive OF or FL techniques. Consequently, the contralateral left quadrant was allocated to the other group. The following treatments were performed:
OF (control group; n=28 sides/105 teeth): An internal beveled incision was performed at the buccal aspect of the involved teeth. Afterwards, a sulcular incision was completed to allow gingival tissue removal. A full-thickness mucoperiosteal flap was reflected to remove and remodel bone tissue by means of surgical chisels, as necessary, until a 3mm distance was achieved between the bone crest and the cement-enamel junction (CEJ). The exposed root surfaces were carefully planed with curettes. Interrupted sutures were performed at the papilla to allow GM stabilization in the CEJ position.
FL (test group; n=28 sides/105 teeth): Internal beveled and sulcular incisions and gingival tissue removal were performed as above described for the control group, replacing the GM in the CEJ position. However, the alveolar bone was removed and remodeled, as necessary, using micro chisels, via incisions, without flap elevation. The root surfaces were also carefully planed via incisions. The required distance of 3mm between the bone crest and the CEJ was checked by inserting a periodontal probe into the incision. Sutures were not performed.
After both procedures, canines and central incisors should be at the same length, and the lateral incisor should be 1mm shorter. The same periodontist (F.R.V.) performed all surgeries. Chlorhexidine gluconate mouthwash (0.12%) was prescribed 2x/day for 2 weeks. Analgesics were prescribed to control possible postoperative discomfort. The sutures of control quadrants were removed after 7 days. The surgical time for test and control, starting after anesthesia, was computed.
Examiner calibration
Clinical examinations were performed by one trained examiner (D.H.), calibrated as previously described13. The intra-examiner variability was 0.16 mm for PD and 0.18 mm for CAL. The parameters registered dichotomously (e.g. BoP) were calculated by the Kappa-Light test and the intra-examiner agreement was \> 0.85.
Clinical monitoring
Periodontal measurements were performed using a manual probe . An individual stent of ethylene-vinyl acetate copolymer was used as a reference point for some clinical parameters. Grooves were drilled in the stent to standardize the localization and direction of the probe. The following parameters were assessed on the mesio-buccal, middle-buccal and disto-buccal aspect of all included teeth at baseline, 3, 6 and 12 months post-surgeries: plaque accumulation (PI)12; MB: bleeding up to 15 seconds after probing along GM; BoP: bleeding up to 15 seconds after probing, at least 1 min following MB recording; PD: distance between the GM and the bottom of the gingival sulcus; Relative GM (rGM): distance from a fixed landmark in the stent to the most cervical point of the GM; Relative CAL (rCAL): distance from a fixed landmark in the stent to the bottom of the gingival sulcus; Keratinized gingiva height (KGH): distance from the GM to the mucogingival line. rGM was also assessed immediately after surgery.
The relative bone level (rBL, i.e. the distance between a fixed landmark in the stent and the bone crest \[BC\]) was recorded before and after the surgical procedures.
Enzyme linked immunosorbent assay (ELISA)
To evaluate the effects of surgeries on alveolar bone remodeling, the gingival crevicular fluid (GCF) levels of receptor activator of NF-КB ligand (RANKL) and osteoprotegerin (OPG) were assessed. One site from each quadrant was randomly chosen for GCF sampling at baseline and at 3 months post-surgeries. The GCF sampling protocol and volume measurement were performed as previously described14. The strips were stored at -80ºC for subsequent assays. GCF samples were analyzed by ELISA for soluble RANKL (sRANKL) and OPG using commercially available ELISA kits , according to the manufacturer's recommendations14. Results were reported as total amount (pg) and concentration (pg/µl of GCF) of protein.
Cone-beam computed tomography (CBCT)
To evaluate the hard and soft tissues of the buccal periodontal apparatus, a soft tissue CBCT (ST-CBCT) was performed as previously described15. At baseline, the CBCT scans were taken with an iCATǁ and the images (sections of 1.0 mm apart) were acquired using suitable software. Briefly, the individuals were submitted to a regular CBCT scan, wearing a plastic lip retractor in an inverted position and retracting their tongues toward the floor of the mouth. These procedures prevented the interference of the soft tissues of the lips, cheeks and tongue in the gingival tissue. The following measurements were obtained in the buccal aspect of the middle image section of each tooth with a digital caliper rule: bone thickness (BT, the width of the buccal bone at a distance of 3 mm apical to the bone crest), CEJ to BC distance (CEJ-BC), gingival thickness (GT, the width of the buccal gingival tissue at a distance of 3 mm apical to the bone crest), GM to BC distance (GM-BC) and CEJ to GM distance (CEJ-GM).
Patient perceptions
Patient perceptions regarding morbidity and esthetic satisfaction were evaluated with a questionnaire administered by an assistant (TSM). The questionnaire was obtained upon completion of the procedure (pain) and at 7 days (pain/discomfort, swelling, hematoma, esthetical appearance) and 6 months post-surgery (esthetical appearance). Responses were quantified with a visual analogue scale (VAS) of 100 mm in which 0 indicated 'no' and 100 'plenty.' In addition, at 7 days and 6 months post-surgery, a questionnaire recorded patient satisfaction regarding the type of treatment, in terms of esthetical appearance by selecting one of the following choices: totally satisfied, partially satisfied or unsatisfied.
Statistical Analysis
To validate the clinical comparisons of this paper, a post hoc power calculation was performed based on differences of 0.5mm in GM between groups at 12 months post-surgery. Since there is no previous study comparing OF and FL techniques, we established that a difference of 0.5mm in GM between both approaches could be a relevant clinical parameter to perform the power size calculation. In addition, a standard deviation (SD) of 0.5 mm was determined based on the observed SD of the difference in GM changes between groups at 12 months, considering all buccal sites (interproximal- plus middle-buccal sites). Based on these data, it was determined that 16 subjects per group would be necessary to provide an 80% power with an alpha of 0.05. Since 28 subjects met the inclusion criteria, these were all included in the study. The power calculation took into account the split-mouth design16.
Data were examined for normality by the Shapiro-Wilk test. The data that did not achieve normality were analyzed using non-parametric methods. The mean percentage of sites with visible plaque accumulation, MB, BoP and the mean PD, rGM, rCAL, KGH and rBL were computed, separately, for interproximal- and middle-buccal sites of control and test sides. Interproximal sites included the mean of mesio- and disto-buccal measurements. The changes in the rGM from baseline to immediately after surgery, 3, 6 and 12 months and in rBL from baseline to immediately after surgery were calculated for both groups. Clinical differences between groups were compared using the paired Student's t-test. Repeated measures ANOVA and Tukey test were employed to detect differences within each group among time-points. The Mann-Whitney test was used to evaluate VAS scores of patient's perceptions. The satisfaction in terms of esthetical appearance was compared by the χ2test. RANKL and OPG differences between groups and time-points were compared using the Wilcoxon test.
A model of multiple linear regression (MLR) analysis was performed to estimate the association between the dimensions of the soft and bone tissues obtained by ST-CBCT, the treatment modalities and the changes in the GM. The outcome variable in this model was the nonappearance of creeping attachment of the GM in a coronal direction, at 12 months, from the position defined immediately after surgery (yes/no). The predictor variables included the surgical modalities and the tomography measurements. The level of significance was set at 5% for all analyses.
