Stability of maxillary anterior crowding treatment

Methods: The experimental sample consisted of 70 patients of both sex with an initial Class I and Class II maloclusion and treated with first premolar extractions. The initial mean age was 13,08 years. Dental casts’ measurements were obtained at three stages (pretreatment, posttreatment and posttreatment of 9 years on average) and the variables assessed were Little Irregularity Index, maxillary arch length and intercanine. Pearson correlation coefficient was used to know if some studied variable would have influence on the crowding in the three stages (LII1, LII2, LII3) and in each linear displacement of the Little irregularity index (A, B, C, D, E) in the initial and post-retention phases.


INTRODUCTION
In the last decades many patients have been seeking orthodontic treatment for esthetics reasons. Orthodontic treatment can improve facial esthetics as well as the occlusion, but long-term stability of the aligned teeth is highly variable and unpredictable. A wide variability of long-term results may be related to the amount of pretreatment crowding, treatment plan, patient's age and cooperation during and after treatment. 22 For years, many papers on alignment stability researched mandibular anterior crowding, probably because relapse of these teeth are greater than that of the maxillary anterior teeth. 4 More recently, patient's expectation have been considered, 2,9,31 and for the patient, the alignment of the maxillary anterior teeth is especially important, since those teeth are the first to be shown on a smile. 10,22 Since there is greater concern on the esthetics, a small relapse could be a problem. This new scenario obligates the orthodontists to seek out knowledge in order to inform their patients about this problem and to control the risk factors during and after orthodontic treatment.
Most studies have shown that crowding relapse appears to be multifactorial. 10,15 The amount of initial crowding, the arch length, intercanine distance are the most studied factors. There is a consensus about the teeth's tendency to return toward their original position, 1,5,12,14,22 but these studies are, in general, based on arches changes during the orthodontic treatment.
Therefore, in order to seek out for more knowledge on long term stability of maxillary anterior teeth on the orthodontic treatment, this paper evaluates the relapse of the maxillary anterior teeth in cases treated with premolar extractions and their tendency to return toward their original position.

MATERIAL AND METHODS
The sample was selected from the patient records treated in the Department of Orthodontics at Bauru Dental School, University of São Paulo and in the ACOPEN (Assessoria e Consultoria em Ortodontia, Pesquisa e Ensino). To minimize the bias, the sample was selected by inclusion criteria based on the literature 4,10,12,17,19,25,29,30 and on the objective of this study. The selection criteria were patients with all permanent teeth erupted up to the first molars and under 15 years of age at pretreatment (T 1 ); no supernumerary teeth or tooth agenesis; no fiberotomy or interproximal stripping as part of the treatment plan; Class I or Class II Division 1 malocclusion of, at least, three-forths of a full step Class II molar relationship, no anterior open bite or crossbite and more than 3 mm of crowding on Little's irregularity index 9 in the mandible (LIIMx). The pretreatment (T 1 ), posttreatment (T 2 ) and post-retention (T 3 ) dental casts had to be in good condition to be evaluated. All patients were treated with fixed edgewise appliances and they not underwent rapid maxillary expansion. The patients also had a maximum peer assessment rating (PAR) score of 5 and passive lip seal at posttreatment (T 2 ). Retention included a maxillary Hawley plate, used it 24 hours per day, for 6 months minimal plus 3 months during sleeping, and a bonded lingual canine-to-canine retainer in the mandibular arch. The post-retention dental casts (T 3 ) had to be at least 5 years after treatment. The presence of third molars was not part of the inclusion criteria because there is no common sense that have shown these teeth really interfere in anterior crowding relapse. 3,13,20,28 Therefore, the sample comprised 70 patients, 210 pairs of dental casts (pretreatment, posttreatment and post-retention).

Variable collection method
All dental cast measurements were made with a 0.01 mm precision digital caliper and capacity of 150 mm (Mitutoyo America, Aurora, Ill) by the same examiner. The examiner was blinded in relation to which group did the cast belong to in order to minimize the bias.

The assessed variables: Little's irregularity index (LII)
The index used to evaluate the mandibular anterior crowding was proposed by Little 18 and named after him. The Little's irregularity index was used because of its great reproductive, precision and it is used in the most studied about crowding and relapse. The variables described by this irregularity index were adapted to be evaluated on the maxillary arch. This adaptation was used in previous studies. 6,10,21 In order to measure the amount of crowding a digital caliper was positioned parallel to the occlusal plane. Each linear displacement between the 5 anatomic contact points (A, B, C, D, E) of the anterior teeth were measure (Fig 1). Little's irregularity index comprises the 5 linear displacements added up, which represents the amount of crowding in the anterior teeth.

Arch length (AL)
It is the total distance, in millimeters, from the contact point of the maxillary central incisors to the mesial of the first permanent molars in the right (F) and left side (G) (Fig 2).

Intercanine Width (ICW)
Distance from the cusp tip of the upper canines in millimeters. In cases of cusp wear, the tip was estimated (Fig 2).

Statistical method
All statistical analyses were performed with Statistica software (version 6.0, Statsoft, Tulsa, Oklahoma). Normal distribution was verified by the Kolmogorov-Smirnov test and the results were considered significant when p<0.05. The descriptive analysis found the mean, minimum, maximum, standard deviation, the groups and subgroups in all variables studied in the pretreatment (T 1 ); posttreatment (T 2 ), post-retention (T 3 ) casts, as well as, the difference between the posttreatment and pretreatment phase (T 2 -T 1 ) and the post-retention and posttreatment phase (T 3 -T 2 ). The difference between T 2 and T 1 is considered to be treatment changes and the difference between T 3 and T 2 is posttreatment changes.

Method error
The reliability of this study was evaluated by repeated measures of the variables of 15 patients (all phases) randomly chosen. The examiner had a month of interval between the first and the second measurement.
The systematic and casual error were evaluated for each variable. Systematic errors were evaluated with paired t tests at p<0.05, according to Houston. 16 Casual errors were calculated according to Dahlberg's formula (Se 2 = Sd 2 /2n). 7

Statistical analyses
Many studies in the literature are composed with Class I and Class II malocclusions in the same group. In this study, before the Class I and Class II malocclusion joined the same group the compatibility of  Stability of maxillary anterior crowding treatment original article these two malocclusions was evaluated. The sample was divided into 3 groups: Group 1 (Class I malocclusion treated with 4 first premolar extraction), group 2 (Class II division 1 malocclusion treated with 4 first premolar extraction), group 3 (Class II division 1 malocclusion treated with 2 first maxillary premolar extraction). Intergroup comparisons of all variables were made by one-way analysis of variance (ANOVA). Intergroup sex distribution was evaluated with the chi-square test.
Pearson correlation coefficients were used to assess the relationships between the variables (LI-IMx, AL, ICW, posttreatment time and post-retention time). Another evaluation, was the correlation between each linear displacement of Little's irregularity index (A, B, C, D, E). In that way, it was possible to verify the tendency of the teeth to return toward their original position (A1A3, B1B3, C1C3, D1D3, E1D3). The relapse also was evaluated in each linear displacement from Little's irregularity index by using the scores in the pretreatment (T 1 ) and post-retention (T 3 ) phases. Any contact point that was correctly adjusted (0 mm), in any one of these phases (T 1 or T 3 ), was not considered. All other linear displacements were used. The percentages of teeth that kept the same labiolingual direction at T 1 and T 3 were calculated.
The last comparison regarded the severity of the pretreatment crowding. The whole sample was divided in 2 groups (A and B). Group A comprised patients with LIIMx scores less than 7 mm, or minimal and moderate irregularity (19 patients), while group B had LIIMx scores equal to or greater than 7 mm, or severe and very severe irregularity (51 patients). The ratio between the post-retention changes (LIIMx3-2) and the correction amounts (LIIMx2-1) was called the relapse percentage. The absolute score of the correction amounts was used. The posttreatment changes that had negative scores (a greater alignment of the contact point) was considered to be zero.

Method error
Variables showed casual error smaller than 1 mm. Among all 30 variables, only width D at T 2 showed a significant systematic error (96% precision).

Statistical analises
The table 1 and 2 show the compatibility between the 3 groups regarding variables (Tab 1) and gender ( Tab 2).  The table 3 shows the mean, minimum, maximum, standard deviation and total sample. Pearson correlation coefficients were used to assess the relationships of LIIMx scores at different stages with the other variables (Tab 4). There was a positive correlation between LIIMx1 x LIIMx3, LIIMx1 x LI-IMx3-2 and LIIMx2 x LIIMx3. There was a negative correlation between LIIMx2-1 x LIIMx3-2. Table 5 shows the correlation between the linear distance in the post-retention (T 3 ) and pretreatment (T 1 ) phase in B, C and D.
The sample consisted of 70 patients, each patient had 5 anatomic contact points described by Little 18 (Fig 1); totalling of 350 contact points. In the pretreatment phase, there were 309 linear displacements for labiolingual direction. The amount of linear displacement in post-retention phase was verified at the same contact points of the 309 linear displacements in pretreatment. The total of linear displacement in post-retention phase was 184. These 184 linear displacements were used to evaluate the tendency that a tooth had to return toward its original position. A total of 142 anatomic contact points in the pretreatment phase had the same labiolingual direction in the post-retention phase. These results have shown that 77% of the labiolingual linear displacement had the tendency to return toward their original position.
The total sample was divided in group A (LI-IMx1 < 7 mm) and group B (LIIMx1 > 7 mm) in order to evaluate the pretreatment severity crowding with the relapse percentage. The paired t test did not show a significant difference among these variables (Tab 6).

DISCUSSION
Groups 1, 2, 3 were compatible regarding all variables (Tab 1 and 2), therefore the total sample (70 patients) could be evaluated. The maxillary anterior crowding relapse (LIIMx3-2) of the total sample was 1.07 mm (Tab 3). Since the percentage is the easiest way to visualize a result, there was an 11.88% of maxillary anterior crowding relapse. This percentage was the result from the posttreatment changes Quaglio CL, Freitas KMS, Freitas MR, Janson G, Henriques JFC (LIIMx3-2) divided by the treatment changes (LI-IMx2-1) multiplied by 100. Others studies in the orthodontic literature had similar results 10,11,15,21,22 . Our result showed a great stability of the maxillary anterior alignment, 88,12%.
Pearson correlation test was used to assess the relationships of LIIMx scores at different stages with the other variables. There was a significant positive correlation between Little's irregularity index at pretreatment (LIIMx1) and the posttreatment changes (LIIMx3-2). This result shows that pretreatment maxillary anterior crowding is correlated with the posttreatment maxillary anterior crowding (Tab 4), and this is supported by previous studies that have shown that pretreatment maxillary anterior crowding interfere in the crowding relapse. 1,8,24 The posttreatment Little's irregularity index (LIIMx2) has a significant positive correlation with the post-retention Little's irregularity index (LIIMx3). The posttreatment changes (LIIMx3-2) did not show a significant correlation with the amount of posttreatment crowding (LIIMx2) (Tab 4). Both results must be analyzed together because the crowding relapse would be influenced by quality of treatment results (LIIMx2) only if there was a significant correlation between the amounts of posttreatment crowding (LIIMx2) and the posttreatment changes. Many previous studies, even   Stability of maxillary anterior crowding treatment original article though being about mandibular anterior crowding, corroborated with this study. 22,23,24 The positive significant correlation between posttreatment Little's irregularity index (LIIMx2) and post-retention Little's irregularity index (LIIMx3) is that posttreatment crowding is, at least, the same after a long-term evaluation.
Changes during treatment (LIIMx2-1) had a significant correlation with the post-retention changes (LIIM x 3-2) (Tab 4). Despite this negative correlation, it is considered to be a false negative. Since LI-IMx2-1 has a negative sign (posttreatment crowding minus pretreatment crowding) and the variable LI-IMx3-2 has a positive sign (post-retention crowding minus pretreatment crowding); when these two variables are correlated, the negative sign is maintained as a positive correlation result.
The post-retention crowding (LIIMx3) was evaluated with the posttreatment arch length and intercanine width (AL3, ICW3), posttreatment observation and age in all phases. These variables were chosen because the literature presented some studies with significant positive correlation between the post-retention crowding (LIIMx3) and these variables 1,27,30 (Tab 4). None of these variables was significantly correlated.
The Pearson correlation coefficients were used to assess the relationships between linear displacements of the anatomic contact points at T 1 and T 3 . The percentages of teeth that kept the same labiolingual direction at T 1 and T 3 were calculated. There were significant positive correlations among linear displacements of anatomic contact point B (mesial surface of maxillary right lateral incisor with distal surface of maxillary right central incisor), contact point C (mesial surface of maxillary right central incisor with distal surface of maxillary left central incisor), and contact point D (mesial surface of maxillary left lateral incisor with distal maxillary left central incisor) at T 1 and T 3 (Tab 5). These small proximal surfaces could cause a weak contact point which could increase susceptibility of misalignment over the years. The literature has not evaluated the recurrence of crowding the way our study has, in each region of the Little Index (A, B, C, D and E), so our study should not be compared directly with any study published.  *p< 0,05%.