Introduction
The objectives of this 2-arm parallel trial were to compare the numbers of failures of mandibular fixed retainers bonded with indirect and direct methods and to investigate the posttreatment changes 2 years after placement.
Methods
Sixty-four consecutive patients from the postgraduate orthodontic clinic of the University of Geneva in Switzerland were randomly allocated to either an indirect or a traditional direct bonding procedure of a mandibular fixed retainer at the end of their orthodontic treatment (T0). Eligibility criteria were the presence of the 4 mandibular incisors and the 2 mandibular canines, and no active caries, restorations, fractures, or periodontal disease of these teeth. The patients were randomized in blocks of 4 (using an online randomization service) with allocation concealment secured by contacting the sequence generator for assignment. The patients were recalled 12 months and 24 months (T3) after retainer bonding. The main outcome was any first-time failure of retainers (ie, at least 1 composite pad debonded or fractured); unexpected posttreatment changes of the mandibular incisors and canines were a secondary outcome. Impressions and lateral cephalograms were taken at T0 and T3: changes in mandibular intercanine and interpremolar distances and mandibular incisor inclination were assessed. Blinding was applicable for outcome assessment only. The chi-square test and Cox regression were used to compare the survival rates of the retainers bonded with direct and indirect methods. Paired t tests were used to assess differences in intercanine and interpremolar distances and mandibular incisor inclination at T0 and T3. Significance was set at P <0.05.
Results
Sixty-four patients were randomized in a 1:1 ratio. One patient dropped out at baseline, and 3 patients did not reach the T3 recall. In 24 of 60 (40%) patients, the fixed retainer failed within 2 years: 13 of 30 (43%) in the indirect bonding group and 11 of 30 (37%) in the direct bonding group (log-rank test, P = 0.64). The hazard ratio was 1.26 (95% confidence interval, 0.56-2.81; P = 0.58). Bond failures occurred mainly during the first year. There were no clinically significant changes in mandibular intercanine distance, interpremolar distance, and incisor inclination between T0 and T3, or between groups. In 5 patients (17%), all in the direct bonding group, unexpected posttreatment changes, systematically consisting in a lingual inclination of the mandibular left canine, were observed. In 1 patient (3%), the change was considered clinically severe. No other serious harm was observed.
Conclusions
There was no difference in the risks of failure between mandibular retainers bonded with direct and indirect methods. Bonded retainers are effective in maintaining intercanine and interpremolar distances. There seem to be fewer unexpected posttreatment changes with retainers bonded with the indirect compared with the direct method.
Registration
The trial was not registered.
Protocol
The protocol was not published before trial commencement.
Funding
No funding or conflict of interest to be declared.
Highlights
- •
Lingual mandibular retainers bonded with direct and indirect methods were assessed.
- •
Two years after bonding, risks of failure did not differ.
- •
Bonded lingual retainers maintained intercanine and interpremolar distances.
- •
Fewer unexpected changes occurred with the indirect method compared with the direct.
- •
Severe unexpected changes were observed in 1 patient of 60 (1.6%).
The use of a lingual fixed retainer is crucial for maintaining stability in the mandibular anterior region of most patients after orthodontic treatment. To ensure this stability, orthodontists need fixed retainers with limited risks of bond failure and posttreatment changes related to distortion or residual activity of the wire. A mandibular fixed retainer from canine to canine is the most used and accepted method for long-term retention. The first direct bonded retainer was introduced by Knierim. In 1977, Zachrisson introduced the multistranded bonded lingual retainer, which, although varying in wire types, diameters, and bonding procedures, has become the gold standard.
Although a rigid mandibular retainer bonded on the canines only is described as solid and easy to place, it does not prevent relapse of the incisors. Renkema et al found an increase in irregularity index with a mandibular fixed retainer bonded on the canines only within a 5-year posttreatment period.
Retainers can be bonded with a direct or an indirect bonding method. The direct bonding method requires the construction of a retainer on the patient’s cast, subsequently bonded and light-cured in the mouth, after using a transfer key to keep the retainer wire in the right position. The indirect bonding method relies on the preparation of composite pads on the patient’s cast to be bonded using a transfer tray covering the retainer and composite pads. Indirect bonding of a mandibular fixed retainer is a clinically faster procedure than direct bonding.
A recent systematic review about fixed retainers reported a wide variety in protocols, differing in number of teeth bonded (all 6 anterior teeth vs canines only), bonding material, type and dimension of wires. The authors also reported wide variations in the risks of bond failure, varying from 3.5% to 53% for metallic retainers and from 11% to 51% for glass fiber reinforced retainers. It was suggested that most fixed retainer failures occur during the first 3 to 6 months, whereas the probability of failure significantly drops after a year. Most studies evaluating the failure risk of mandibular fixed retainers are retrospective, without clinical information concerning wire preparation or bonding method, and with clinical procedures performed by different operators, resulting in wide variations in so-called operator sensitivity. A few long-term prospective studies have compared the survival rates of different designs of mandibular fixed retainers with a direct bonding method. However, only 2 prospective studies have investigated the influence of direct vs indirect bonding methods on survival rates, with follow-ups limited to 6 months. Therefore, long-term prospective studies are needed to evaluate whether there is an influence of the bonding method (direct or indirect) on the failure rate and to identify the most effective mandibular fixed retainer in the long run.
Regarding posttreatment stability, unexpected movements of anterior teeth have been reported and are related to an active component of the wire or an operator-induced elastic deflection of the wire during bonding of a fixed retainer. It could be hypothesized that the indirect bonding method, where the wire is locked in a fixed position with the composite pads already prepared, could offer the advantage of bonding the wire absolutely passively, therefore preventing unexpected posttreatment movements. It has not been previously investigated whether direct vs indirect bonding procedures could induce different outcomes in terms of unexpected posttreatment changes.
In general, stability after orthodontic treatment is variable, but the highest posttreatment relapse within 10 years was found to occur during the first 2 years after debonding. Therefore, a follow-up period of 2 years seems appropriate for a long-term assessment of mandibular fixed retainers.
Specific objectives and hypotheses
The aims of this study were (1) to assess the numbers of failures of direct and indirect bonded retainers at 2 years and to determine the time frame associated with the highest risk of debonding, and (2) to investigate the posttreatment changes (failure of stability) 2 years after bonding of mandibular fixed retainers, with either the indirect or direct bonding method.
Our hypotheses were the following: (1) the numbers of failures 2 years after bonding mandibular fixed retainers with indirect or direct bonding are similar, and most debondings occur within the first year; and (2) there are fewer unexpected posttreatment changes (failure of stability) with the indirect bonding method compared with the direct bonding method.
Material and methods
Trial design and any changes after trial commencement
This study was a single-center 2-arm parallel randomized controlled trial with a 1:1 allocation ratio. No changes occurred after trial commencement.
Participants, eligibility criteria, and settings
The sample of this randomized controlled trial was previously described by Bovali et al. Consecutive patients who had completed their orthodontic treatment had a mandibular fixed retainer placed by the same operator (E.B.), with either a direct or an indirect bonding method, between September 2012 and June 2013 in the postgraduate orthodontic clinic of the University of Geneva in Switzerland. The inclusion criteria consisted of presence of the 4 mandibular incisors and the 2 canines, and no active caries, restorations, fractures, or periodontal disease of these teeth. Each patient (or a parent in case of a child) was asked to sign an informed consent before the bonding of the mandibular fixed retainer. The protocol of this study was approved by the Ethical Committee of Research of the Human Being at the University of Geneva (reference number 12-198/psy 12-020).
Interventions
All retainers (0.0215-in multistrand stainless steel wire) were constructed by the same laboratory technician and passively adapted to the lingual surfaces of the incisors and canines on the plaster models. The 2 bonding methods (direct and indirect) were described by Bovali et al.
The patients were initially recalled after 1, 2, 4, and 6 months, and impressions were taken after 6 months (T1). Subsequently, all patients had 2 additional retention visits at 1 year (T2) and 2 years (T3) after debonding and retainer placement, respectively. The following steps were taken by an investigator (F.E.) at the T2 or T3 retention visits: (1) clinical examination (T2 and T3), with reports of debonding of the mandibular fixed retainer (fixed retainers having at least 1 composite pad debonded were considered failures), and reports of change in torque or rotations of the mandibular incisors and canines; (2) alginate impression of the mandibular arch (T3 only); and (3) lateral cephalogram (T3 only).
Plaster dental casts were produced from the impressions. All measurements were performed by the same investigator (F.E.).
The following dental casts measurements were made.
- 1.
Intercanine and interpremolar (first premolar) distances at bonding of the retainer (T0) and T3 were measured with a digital caliper (Fino, Bad Bocklet, Germany) with an accuracy of 0.01 mm. The intercanine and interpremolar distances were measured from the center of the cusp of the mandibular left canine or premolar (buccal cusp) to the center of the cusp of the mandibular right canine or premolar (buccal cusp). In case of abrasion of a cusp, the center of the abrasion facet was used instead of the center of the cusp.
- 2.
Changes in torque or rotations of the mandibular incisors and canines observed between T0 and T3 were recorded.
The following cephalometric measurements were made. The angle between the mandibular incisor axis and the mandibular plane (incisor inclination) was measured at T0 and T3 on the lateral cephalograms, using OnyxCeph 2D Pro (version 3.1.58[75]; Image Instruments, Chemnitz, Germany).
Outcomes (primary and secondary) and any changes after trial commencement
The primary outcome was the long-term (2 years) observation of retainer failures with the 2 bonding methods. The secondary outcome was the observation of unexpected posttreatment changes in both groups. No changes were made after trial commencement.
Sample size calculation
The sample size calculation was previously described and was performed according to the primary outcome of the previous report: retainer bonding time.
Interim analyses and stopping guidelines
Not applicable.
Randomization
The method of randomization was previously described. Briefly, the patients were randomized in blocks of 4 (using an online randomization service) to either the direct or the indirect bonding method. Allocation concealment was secured by contacting the sequence generator for assignment.
Blinding
Blinding of either patient or operator (E.B.) was evidently not possible during the bonding procedure. Blinding of the investigator (F.E.) for the clinical assessment of the models and the cephalometric measurements was ensured by assigning a number to each patient to anonymize the data and make sure that the assessor was unaware of the patient’s bonding method.
Statistical analysis
The chi-square test was used to compare the 2-year survival rates of the retainers bonded with direct and indirect bonding. The log-rank test was used to compare the Kaplan-Meier survival curves of both bonding methods, and Cox regression modeling was used to calculate hazard ratios and 95% confidence intervals (CI). Independent t tests were used to compare the changes in intercanine and interpremolar distances and the changes in inclination of the mandibular incisors between the groups (direct and indirect bonding methods). Paired t tests were used to assess the differences in intercanine and interpremolar distances as well as mandibular incisor inclination in the groups at T0 and T3. A chi-square test was applied to analyze the influence of an extraction or a nonextraction treatment approach on the increase or decrease in intercanine distance. All analyses were conducted using SPSS software (version 22; IBM, Armonk, NY).
Error of the method
To determine the intraobserver reliability of the measurements (intercanine and interpremolar distances as well as mandibular incisor inclination), 30 randomly selected dental casts and lateral cephalograms were remeasured by the same operator more than 2 months after the first assessment. Paired t tests were applied between both series of measurements to assess possible systematic errors. The intraclass correlation coefficient and the Dahlberg formula (D = √ [Σd 2 /2N]), where d is the difference between the first and second measurements, and N is the sample size that was remeasured, were calculated to assess random errors.
Material and methods
Trial design and any changes after trial commencement
This study was a single-center 2-arm parallel randomized controlled trial with a 1:1 allocation ratio. No changes occurred after trial commencement.
Participants, eligibility criteria, and settings
The sample of this randomized controlled trial was previously described by Bovali et al. Consecutive patients who had completed their orthodontic treatment had a mandibular fixed retainer placed by the same operator (E.B.), with either a direct or an indirect bonding method, between September 2012 and June 2013 in the postgraduate orthodontic clinic of the University of Geneva in Switzerland. The inclusion criteria consisted of presence of the 4 mandibular incisors and the 2 canines, and no active caries, restorations, fractures, or periodontal disease of these teeth. Each patient (or a parent in case of a child) was asked to sign an informed consent before the bonding of the mandibular fixed retainer. The protocol of this study was approved by the Ethical Committee of Research of the Human Being at the University of Geneva (reference number 12-198/psy 12-020).
Interventions
All retainers (0.0215-in multistrand stainless steel wire) were constructed by the same laboratory technician and passively adapted to the lingual surfaces of the incisors and canines on the plaster models. The 2 bonding methods (direct and indirect) were described by Bovali et al.
The patients were initially recalled after 1, 2, 4, and 6 months, and impressions were taken after 6 months (T1). Subsequently, all patients had 2 additional retention visits at 1 year (T2) and 2 years (T3) after debonding and retainer placement, respectively. The following steps were taken by an investigator (F.E.) at the T2 or T3 retention visits: (1) clinical examination (T2 and T3), with reports of debonding of the mandibular fixed retainer (fixed retainers having at least 1 composite pad debonded were considered failures), and reports of change in torque or rotations of the mandibular incisors and canines; (2) alginate impression of the mandibular arch (T3 only); and (3) lateral cephalogram (T3 only).
Plaster dental casts were produced from the impressions. All measurements were performed by the same investigator (F.E.).
The following dental casts measurements were made.
- 1.
Intercanine and interpremolar (first premolar) distances at bonding of the retainer (T0) and T3 were measured with a digital caliper (Fino, Bad Bocklet, Germany) with an accuracy of 0.01 mm. The intercanine and interpremolar distances were measured from the center of the cusp of the mandibular left canine or premolar (buccal cusp) to the center of the cusp of the mandibular right canine or premolar (buccal cusp). In case of abrasion of a cusp, the center of the abrasion facet was used instead of the center of the cusp.
- 2.
Changes in torque or rotations of the mandibular incisors and canines observed between T0 and T3 were recorded.
The following cephalometric measurements were made. The angle between the mandibular incisor axis and the mandibular plane (incisor inclination) was measured at T0 and T3 on the lateral cephalograms, using OnyxCeph 2D Pro (version 3.1.58[75]; Image Instruments, Chemnitz, Germany).
Outcomes (primary and secondary) and any changes after trial commencement
The primary outcome was the long-term (2 years) observation of retainer failures with the 2 bonding methods. The secondary outcome was the observation of unexpected posttreatment changes in both groups. No changes were made after trial commencement.
Sample size calculation
The sample size calculation was previously described and was performed according to the primary outcome of the previous report: retainer bonding time.
Interim analyses and stopping guidelines
Not applicable.
Randomization
The method of randomization was previously described. Briefly, the patients were randomized in blocks of 4 (using an online randomization service) to either the direct or the indirect bonding method. Allocation concealment was secured by contacting the sequence generator for assignment.
Blinding
Blinding of either patient or operator (E.B.) was evidently not possible during the bonding procedure. Blinding of the investigator (F.E.) for the clinical assessment of the models and the cephalometric measurements was ensured by assigning a number to each patient to anonymize the data and make sure that the assessor was unaware of the patient’s bonding method.
Statistical analysis
The chi-square test was used to compare the 2-year survival rates of the retainers bonded with direct and indirect bonding. The log-rank test was used to compare the Kaplan-Meier survival curves of both bonding methods, and Cox regression modeling was used to calculate hazard ratios and 95% confidence intervals (CI). Independent t tests were used to compare the changes in intercanine and interpremolar distances and the changes in inclination of the mandibular incisors between the groups (direct and indirect bonding methods). Paired t tests were used to assess the differences in intercanine and interpremolar distances as well as mandibular incisor inclination in the groups at T0 and T3. A chi-square test was applied to analyze the influence of an extraction or a nonextraction treatment approach on the increase or decrease in intercanine distance. All analyses were conducted using SPSS software (version 22; IBM, Armonk, NY).
Error of the method
To determine the intraobserver reliability of the measurements (intercanine and interpremolar distances as well as mandibular incisor inclination), 30 randomly selected dental casts and lateral cephalograms were remeasured by the same operator more than 2 months after the first assessment. Paired t tests were applied between both series of measurements to assess possible systematic errors. The intraclass correlation coefficient and the Dahlberg formula (D = √ [Σd 2 /2N]), where d is the difference between the first and second measurements, and N is the sample size that was remeasured, were calculated to assess random errors.
Results
Participant flow
Initially, 66 consecutive patients fulfilled the inclusion criteria. Two patients declined to participate; therefore, 64 patients were randomized in blocks of 4 to either the direct bonding or the indirect bonding method. One patient in the direct group was considered a dropout from baseline, since he did not attend any retention visit. Three other patients were lost to follow-up: 2 in the indirect group and 1 in the direct group failed to complete the trial because they did not attend the 2-year (T3) retention visit. This resulted in the final analysis of 30 patients in the indirect group and 30 patients in the direct group (CONSORT flow diagram; Fig 1 ).
