Three-dimensional microanalysis of tooth movement during the first 6 months of orthodontic double retention

Introduction

This study performed a 3-dimensional analysis of tooth movement during orthodontic retention to assess the effectiveness of double retention (fixed and removable) in preventing undesired tooth movement.

Methods

One hundred randomly selected patients were included at the initiation of double orthodontic retention with fixed retainers and vacuum-formed splints (recommended to be worn 22 h/d) in both arches. Intraoral scans were performed directly (T0), 1 month (n = 88), 3 months (T2) (n = 78), and 6 months (T3) (n = 66) after retainer bonding. Nine reference points were marked on each tooth in every patient. Subsequent scans were superimposed, and point displacement was calculated. Statistical analysis was performed using the R statistical software (version 4.2.2; R Core Team, Vienna, Austria).

Results

Sample size calculation determined at least 55 patients were needed. The total dropout between T0 and T3 was 34 patients (did not show up for appointment). The median absolute displacement value of a single point between T0 and T3 was 0.015 mm. The most stable teeth were mandibular central incisors, whereas the least stable were mandibular molars. Most tooth displacements occurred between T0 and T2, then slowed down significantly.

Conclusions

Double orthodontic retention prevents major tooth displacements in most patients during the first 6 months of retention; however, larger, unpredictable single-tooth displacement may occur in individual patients.

Highlights

•

Double orthodontic retention prevents major tooth displacement in most patients.
•

Under double retention teeth, positions are stable with proper patient cooperation.
•

Most tooth displacement occurs in the first 3 months and then slows down.
•

The most stable teeth were mandibular central incisors, and the least were mandibular molars.
•

Patient engagement during the retention phase of treatment decreases with time.

In recent years, orthodontic treatment has been extremely popular, and after debonding, the patients want to maintain the treatment results. Orthodontic retention is defined as maintaining the optimal esthetic and functional teeth positions achieved during active treatment. It should be considered an important, integral phase of treatment. The issue of optimal retention procedures has not been resolved regarding what makes clinicians develop their retention protocols as they gain experience. European orthodontists most often follow the protocol of double retention consisting of a bonded fixed retainer and a removable retention appliance with regular follow-up. ^, The most popular retainers are made of stainless steel braided rectangular wire or a gold chain bonded with a flowable composite from canine to canine for fixed retention combined with vacuum-formed splints or Hawley retainer for removable retention.

The treatment result is known to be unstable for many reasons (including the fact that tissues around the teeth need a considerable amount of time to adjust to their new positions). The most used index to assess relapse is Little’s index, proposed in 1975. It is based on irregularity assessment as the sum of the distances between the anatomic contact points of the mandibular anterior teeth on the plaster model. This method is 1 dimensional, imprecise, and limited to mandibular anterior teeth. Moreover, it requires taking an impression, and thus, it is time-consuming. Attempts have been made to calculate this index on standard intraoral photographs; however, it requires proper training and has proved challenging. Despite the use of modern dentistry technologies, no new retention monitoring methods have been developed. The most important publications are often concerned with the anterior segment only, considering only the esthetic aspect while ignoring the shape of the whole dental arch and occlusion. Monitoring the retention process is based on clinical experience and is purely observational in current clinical practice.

Modern 3-dimensional (3D) technologies opened up completely new possibilities in dentistry; the latest generation scanners provide higher accuracy than traditional impressions. Superimposition of 3D models in external software allows the identification of undesired changes and performs 3D objective measurements of tooth movement. ^, Recently, 2 case reports ^, and 1 indirect 3D analysis referred to the use of round flexible spiral wire, which is considered to cause adverse tooth movement. Direct intraoral scanning ensures a higher precision than scanning polysulfide impressions or stone casts in a laboratory. However, no studies have analyzed direct whole-mouth 3D model superimposition with double retention using stainless steel braided wire and a vacuum-formed removable retainer. This is the first study to assess direct 3D intraoral scans for tooth displacement during orthodontic retention. Therefore, this study (1) performed a 3D analysis of tooth movement during orthodontic retention and (2) assessed the effectiveness of double retention in preventing undesired tooth movement.

Material and methods

The study protocol has been registered on clinicaltrials.gov (no. NCT05626335).

The study has been exempt from the approval of the local bioethical committee (decision reference no. KB-012/74/10/2020/Z). All the procedures were in accordance with the Helsinki Declaration of 1975 and its further amendments.

Study group enrollment

The study group comprised 300 adult patients (aged 18-41 years) who completed fixed orthodontic treatment between January 1, 2021 and September 9, 2022. The inclusion criteria were (1) nonextraction treatment, (2) nongrowing patients, (3) normal occlusion after treatment (Class I occlusion), (4) perfect dental alignment after active treatment, (5) normal overjet and overbite, and (6) double retention including fixed retainer bonded from canine to canine in both dental arches and vacuum-formed removable splint

The exclusion criteria were (1) fixed orthodontic treatment in 1 dental arch, (2) hypodontia, (3) extraction, (4) patient treated with orthognathic surgery, (5) imperfect treatment results because of treatment cessation on patient demand or health issues, and (6) craniofacial disorders.

Clinical procedures

For all patients meeting these criteria, every third patient was randomly selected and invited to participate in the study. A written informed consent was obtained from every participant, and 100 patients agreed to participate. All fixed retainers were made of stainless steel braided rectangular wire (Bond-a-braid; Reliance Orthodontic Products, Itasca, Ill) and bonded using a flowable composite for fixed retention (Ortho Connect Flow; GC Orthodontics, Yokohama, Japan) by the same experienced clinician (M.J.). Subsequently, impressions were made for vacuum-formed splints (Duran 1.0; Scheu-Dental GmbH, Iserlohn, Germany). The splints were delivered to the patients on the day of debonding. The effect of each procedure was verified by 2 other independent experienced clinicians (K.G., J.J.-O.).

The patients were recommended to wear removable retainers 22 h/d except for mealtime and oral hygiene during the study period. Patients were instructed to immediately report any failure, especially breakage, loss, or partial debonding of the fixed retainer or vacuum-formed splint breakage. The patients were instructed to apply immediately to the office in case of failure.

Every failure (noticed by the clinician during a scheduled appointment or by the patient), including debonding, fracture, or loss of retention appliance, was noted in the patient’s records.

On the day of debonding, directly after retainer bonding, the teeth were thoroughly dried, and intraoral scans were performed (T0) using 3Shape Trios 4 (3Shape, Copenhagen, Denmark). All patients were invited to repeat the scans after 1 (T1), 3 (T2), and 6 months (T3). The patient’s monitoring program, available as part of the 3Shape Shell software on the scanner, was used during the follow-up visits to communicate with the patient and demonstrate wherever tooth movement occurred. This program independently and automatically segments and superimposes selected scans, allowing for quick movement monitoring.

Analysis of the scans

Nine points were marked on each of the 28 teeth (on labial surfaces) in each patient (252 points per patient).

An example of marking points is shown in Figure 1 .

Then, a superimposition of marked reference points on subsequent scans of each patient was performed automatically by specialized software (GOM Inspect; Zeiss, Braunschweig, Germany) using a best-fit-algorithm ( Fig 2 )—each digital scan was analyzed relative to a reference scan (T0). The algorithm analyzed and compared all available points in a 3D model in the optimization process (in which the sum of all squares of deviations is the smallest). Once the 3D models were matched, measuring the displacement of individual reference points was possible. Thus, the displacement of each reference point was measured in 3D.

Thirty-three randomly selected scans were superimposed by another author 1 week after the initial analysis. Intraclass correlation coefficient regarding positioning points on the teeth has been performed to assess agreement between examinations.

The displacement of each reference point among the day of debonding and subsequent scans (T0-T1, T0-T2, and T0-T3) was analyzed. Each displacement of a reference point outside the initial dental arch (eg, buccal movement or extrusion) was noted as a positive value, whereas each displacement inside the initial dental arch (eg, a lingual movement or intrusion) was noted as a negative value.

The teeth were classified in line with the common anatomic classification of tooth types. The range of tooth movement was calculated as a median of the absolute displacement values of 18 reference points (9 on each tooth of the right and left side). The mean movement of each tooth type was calculated as the absolute median value of displacements of these points. The number of reference points analyzed can be calculated using the following equation: 9 points × 2 teeth (of each side) × 4 scans = 132 points for each tooth type. Then, comparisons were made between subsequent periods: T0-T1, T0-T2, and T0-T3.

Then, tooth movement was analyzed for 6 groups of teeth to compare tooth movement in different parts of the dental arch: (1) mandibular anterior teeth (including central mandibular incisors, lateral mandibular incisors, and mandibular canines), (2) mandibular premolars, (3) mandibular molars, (4) maxillary anterior teeth (including maxillary central incisors, maxillary lateral incisors, and maxillary canines), (5) maxillary premolars, and (6) maxillary molars.

The mean movement of the teeth in each group was calculated as the absolute median value of displacements of 56 (for anterior teeth groups) or 36 (for premolar and molar teeth groups) reference points.

Proper blinding during the clinical procedures was not possible as the appointments took place face to face. However, the superimposition tool (Patient Monitoring, 3Shape) assured objective evaluation of displacements between the beginning of retention and subsequent examinations. The evaluation of displacements was blinded, as the investigator, putting points on the scan meshes, did not take part in the clinical phase, and patient data were anonymized.

Sample size calculation was performed referring to the difference in the displacement, using the following formula for continuous variable :

<SPAN role=presentation tabIndex=0 id=MathJax-Element-1-Frame class=MathJax_Error style="POSITION: relative" data-mathml='n=2(z1−α2+z1−β)2d2′>n=2(z1−α2+z1−β)2d2
n=2(z1−α2+z1−β)2d2

In this formula, n is the number of objects in each group, α is the probability of incurring the error of species I or false positive (of rejecting the true null hypothesis [ie, of finding a significant difference because there is no difference]), β is the probability of incurring the second species error or false negative (of not rejecting the false null hypothesis [ie, of not finding a significant difference given that a difference exists]), and d is the Cohen’s effect size.

Statistical analysis

Descriptive statistics were calculated for each tooth and point displacement and 6 groups of teeth (maxillary anterior teeth, mandibular anterior teeth, maxillary premolars, mandibular premolars, maxillary molars, and mandibular molars) in each of 3 points in time (T1, T2, and T3). The Shapiro-Wilk normality test was performed to assess the distribution of all the variables. Considering the significant nonnormality of the distribution of the variables in the groups, the Kruskal-Wallis test was used to assess the significance of the between-group difference in medians, and the Wilcoxon rank sum test with Bonferroni corrections for multiple testing to determine which pairs of groups are different, and to compare the displacement between the periods. The results were considered statistically significant at P <0.05. The R statistical software package (version 4.2.2; R Core Team, Vienna, Austria) was used for the statistical analyses.

Results

Error study and sample size calculation

The repeatability of the point positioning was high, as the intraclass correlation coefficient in a 2-factor mixed model for absolute agreement of a single measurement yielded a P value of <0.05.

Sample sizes were determined for the probability of a type I error with an α = 0.05 and power at 1 − β = 0.9. Considering multiple comparisons between 6 groups of teeth: maxillary anterior teeth (from canine to canine), maxillary premolars, maxillary molars, mandibular anterior teeth (from canine to canine), mandibular premolars and mandibular molars (K = 15 comparisons), the probability of type I error α was adjusted by Bonferroni correction (α/K instead of α), assuming full detention (28 teeth) in each patient. The essential mean difference was set at 0.2 mm. Based on results by Knauf et al, the standard deviation (SD) was set at 0.5. The resulting standardized effect size (ratio of mean difference to SD) was equal to 0.4, and n ≈ 220 teeth in each group. Molar and premolar groups have 4 teeth per patient; thus, 55 patients were needed. Considering the approximate value of the SD, 20%, the sample of this study (n = 66) was sufficient.

Patient compliance and retention failures

The T0 scan was performed on 100 patients. The T1 scan was performed on 86 patients (14 dropped out [did not apply for the scheduled appointments or refused to cooperate]). The T2 scan was performed on 78 patients (8 dropped out [did not apply for the appointment]). The T3 scan was performed on 66 patients (43 females, 23 males) (12 dropped out) ( Fig 3 ).

During the observation period between T0 and T3, there were 18 failures in 14 patients, including 1 splint breakage, 2 completely debonded, and 16 partially debonded retainers. In 3 patients, the failure occurred twice; in all 3 patients, there was a partially debonded retainer. All the patients came on the day when the failure occurred and had it fixed immediately. In the case of splint breakage, a new splint was provided the next day. The splint was made on the basis of a new impression on the day of the patient’s presentation. More failures were noted in maxillary than in mandibular retainers (14 partially debonded, 2 completely debonded), mainly in lateral incisors and canines. In 2 patients with retainers completely debonded, the patients confessed to having debonded the retainer as it was loose on 1 tooth.

Analysis of mean microchanges in the positions of each tooth

The maximum absolute range of a single-point displacement on any tooth was 2.23 mm and pertained to the middle point of the height of the distal border of the labial surface of the maxillary first molar at T3. The median absolute displacement value of a single reference point was 0.015 mm for all follow-ups. From T0 to T1, the median absolute value was 0.065 mm, whereas from T0 to T2 and T0 to T3, median values were 0.051 mm and 0.042 mm, respectively. This shows that the teeth moved in the first month of retention, and then the mean displacement of the reference points referring to the beginning of retention was reduced in the 2 subsequent months.

Detailed raw data on the displacement of each 252 points from T1, T2, and T3 could not be presented within the paper because of the dimensions of the Excel table (758 rows). The distribution of the mean displacement of each tooth type calculated as the absolute median value of displacements of its 9 reference points is presented in Table I .

Table I

Distribution of absolute values of displacement in subsequent periods

Period	Min	Q1	Median	Q3	Max	Mean ± SD
Mandibular central incisors
T0-T1	0.001	0.030	0.067	0.119	0.557	0.094 ± 0.096
T0-T2	0.000	0.026	0.058	0.112	0.330	0.077 ± 0.068
T0-T3	0.000	0.029	0.059	0.093	0.302	0.067 ±.050
Mandibular lateral incisors
T0-T1	0.000	0.024	0.054	0.103	0.470	0.077 ± 0.080
T0-T2	0.000	0.020	0.038	0.073	0.394	0.056 ± 0.056
T0-T3	0.000	0.014	0.043	0.076	0.262	0.051 ± 0.045
Mandibular canines
T0-T1	0.000	0.023	0.044	0.096	0.336	0.073 ± 0.074
T0-T2	0.001	0.014	0.036	0.067	0.272	0.048 ± 0.048
T0-T3	0.000	0.017	0.032	0.056	0.316	0.041 ± 0.040
Mandibular first premolars
T0-T1	0.002	0.026	0.062	0.109	0.339	0.079 ± 0.067
T0-T2	0.001	0.014	0.032	0.056	0.279	0.045 ± 0.047
T0-T3	0.000	0.013	0.03	0.049	0.170	0.037 ± 0.033
Mandibular second premolars
T0-T1	0.000	0.030	0.086	0.158	0.406	0.107 ± 0.091
T0-T2	0.001	0.021	0.041	0.070	0.476	0.053 ± 0.056
T0-T3	0.000	0.019	0.041	0.077	0.270	0.057 ± 0.051
Mandibular first molars
T0-T1	0.001	0.038	0.082	0.142	0.402	0.102 ± 0.082
T0-T2	0.000	0.028	0.053	0.107	0.436	0.077 ± 0.069
T0-T3	0.001	0.023	0.051	0.088	0.366	0.064 ± 0.055
Mandibular second molars
T0-T1	0.002	0.035	0.094	0.184	0.579	0.120 ± 0.103
T0-T2	0.001	0.036	0.089	0.156	0.559	0.108 ± 0.091
T0-T3	0.000	0.040	0.067	0.113	0.290	0.083 ± 0.064
Maxillary central incisors
T0-T1	0.001	0.033	0.061	0.114	0.982	0.091 ± 0.111
T0-T2	0.000	0.023	0.048	0.091	0.347	0.062 ± 0.056
T0-T3	0.000	0.019	0.038	0.073	0.631	0.056 ± 0.075
Maxillary lateral incisors
T0-T1	0.001	0.030	0.054	0.091	1.137	0.075 ± 0.108
T0-T2	0.000	0.016	0.051	0.08	1.231	0.060 ± 0.057
T0-T3	0.002	0.014	0.034	0.064	1.170	0.046 ± 0.048
Maxillary canines
T0-T1	0.000	0.021	0.049	0.095	0.250	0.066 ± 0.058
T0-T2	0.000	0.017	0.033	0.073	0.354	0.054 ± 0.058
T0-T3	0.001	0.016	0.034	0.053	0.396	0.042 ± 0.045
Maxillary first premolars
T0-T1	0.001	0.026	0.061	0.104	0.380	0.078 ± 0.070
T0-T2	0.001	0.026	0.054	0.094	0.438	0.072 ± 0.069
T0-T3	0.000	0.018	0.038	0.069	0.689	0.056 ± 0.075
Maxillary second premolars
T0-T1	0.000	0.029	0.071	0.136	0.289	0.086 ± 0.066
T0-T2	0.000	0.026	0.052	0.094	0.324	0.069 ± 0.057
T0-T3	0.000	0.019	0.045	0.089	0.341	0.061 ±.053
Maxillary first molars
T0-T1	0.002	0.032	0.081	0.139	0.282	0.093 ± 0.068
T0-T2	0.001	0.016	0.050	0.091	0.316	0.069 ± 0.070
T0-T3	0.000	0.030	0.052	0.090	0.276	0.064 ± 0.048
Maxillary second molars
T0-T1	0.000	0.050	0.106	0.171	0.451	0.122 ± 0.095
T0-T2	0.001	0.023	0.065	0.133	0.458	0.094 ± 0.097
T0-T3	0.001	0.029	0.070	0.129	1.480	0.102 ± 0.150