Stability of molar relationship after non-extraction Class II malocclusion treatment

Methods: The sample comprised 39 subjects (16 females, 23 males) with initial Class II malocclusion treated with no extractions, using fixed appliances. Mean age at the beginning of treatment was 12.94 years, at the end of treatment was 15.14 years and at post-retention stage was 21.18 years. Mean treatment time was 2.19 years and mean time of post-treatment evaluation was 6.12 years. To verify the influence of the severity of initial Class II molar relationship in stability of molar relationship, the sample was divided into two groups, one presenting a 1⁄2-cusp or 3⁄4-cusp Class II molar relationship, and the other with full-cusp Class II molar relationship. In dental casts from initial, final and post-retention stages, molar, first and second premolars and canine relationships were measured. Data obtained were analyzed by dependent ANOVA, Tukey and Pearson’s correlation tests, as well as independent t test between the two groups divided by severity of initial molar relationship.


INTRODUCTION
Class II malocclusion does not self-correct in growing patients. 8,17The Class II skeletal pattern is established early and remains until puberty if no orthodontic intervention is performed. 8o this date, several authors have discussed the relationship of the initial malocclusion characteristics with the effectiveness of orthodontic treatment 22,31,40 and the stability of the corrections obtained. 7,33,43ormally, orthodontic treatment takes a long time and uses complex techniques, usually achieving good results; however, these results may be lost in varying degrees after the removal of appliances and retainers. 38rthodontic relapse includes crowding or spacing of teeth, and loss of overbite, overjet correction, and loss of Class II molar relationship correction.
Orthodontic changes of the position of the first permanent molars have a great tendency to relapse. 24ome authors affirm that Class I molar relationship is more stable compared to others and, over time, the mandibular molar tends to distalize in patients with Class II malocclusion. 17For Uhde, Sadowsky and Be-Gole, 39 changes that occur in molar relationship are always towards Class II relation.The changes are of small magnitude and independent of the type of initial malocclusion and the type of treatment.Other authors suggest that, in the long-term, there is minimal relapse in molar relationship and that changes in incisor position and intercuspation of the posterior teeth are statistically significant, although not considered clinically significant. 12,39he stability of Class II malocclusion has been widely studied, however, few studies that have actually evaluated relapse and stability of the correction of Class II molar relationship in models.The studies are mostly directed to a particular type of appliance or treatment protocol.
In this context, the objective of this study was to evaluate the stability of molar relationship in cases with initial Class II malocclusion, treated orthodontically without extractions, correlating with factors such as the severity of initial Class II molar relationship, treatment, retention and post-retention times.Moreover, the objective was to compare the post retention stability of molar relationship between two groups divided according to the severity of the initial Class II molar relationship.

MATERIAL AND METHODS
For the present retrospective study, the sample was obtained from the records of the Department of Orthodontics, Bauru Dental School, University of São Paulo.
The inclusion criteria comprised the following characteristics: » Angle Class II malocclusion, treated without extractions.» Presence of erupted permanent teeth up to first molars, at the beginning of orthodontic treatment.» Absence of tooth agenesis and supernumerary teeth.» No anomalies in size and / or shape of the teeth.» Absence of rotations of the maxillary and mandibular molars in the initial models, which could influence variable measurement.» Orthodontic treatment with fixed appliances, which may include the use of headgear and / or Class II intermaxillary elastics.» Complete orthodontic records, including study models of pre and post treatment and post retention records.All patients in the sample were Caucasian, of both genders.The study models of the pretreatment (T 1 -initial), posttreatment (T 2 -inal) and post-retention phases (T 3 -post-retention at least 2 years ater the end of treatment) were used.
The sample was composed of a total of 39 patients, 16 females and 23 males.The mean age at pretreatment was 12.94 ± 1.21 years, the mean age at the end of treatment was 15.14 ± 1.38 years, and the mean age at post-retention was 21.18 ± 2.65 years.The mean treatment time was 2.19 ± 0.83 years.The mean posttreatment time evaluation, (between the final and post-retention stages) was 6.30 ± 2.60 years.At the end of active orthodontic treatment all patients used a removable Hawley plate retainer in the maxillary arch and a bonded fixed retainer from canine to canine in the mandibular arch.The retainers were used, on average, 1.62 ± 0.49 years.
To evaluate the inluence of the severity of the initial Class II molar relationship in the stability of molar relationship, the sample was divided into two groups: Group 1, 16 patients presenting half-cusp or ¾-cusp initial Class II molar relationship.The mean age at Stability of molar relationship after non-extraction Class II malocclusion treatment original article pretreatment was 13.27 ± 1.11 years, the mean age at the end of treatment was 15.10 ± 1.35 years and the mean age at post-retention was 20.18 ± 2.03 years.The mean treatment time was 1.83 ± 0.49 years, the mean retention time was 1.58 ± 0.55 years and the mean time between posttreatment and post-retention phases was 5.29 ± 1.57 years.Group 2 consisted of 23 patients with a complete initial Class II molar relationship.The mean age at pretreatment was 12.71 ± 1.25 years, the mean age at the end of treatment was 15.17 ± 1.42 years and the mean age at post-retention was 21.87 ± 2.84 years.The mean treatment time was 2.45 ± 0.94 years, the mean retention time was 1.69 ± 0.46 years and the mean time between posttreatment and post-retention phases was 6.70 ± of 2.46 years.
The orthodontic records of the selected sample were used to obtain some relevant data for this work.Clinical records were examined for therapeutic procedures at pre and posttreatment, and the posttreatment followups.The date of removal of retainers was also observed.These data, together with the patient's date of birth, allowed accurate determination of the total time of treatment, control, posttreatment, post-retention, retention time and age of the patients in the studied phases.
The study models concerning initial, inal and post-retention stages of each patient were evaluated.Study models were photographed with a D-80 camera, with 105 mm close-up lens (AF-S VR Micro Nikkor 105 mm f/2.8GIF-ED, Nikon Corporation, Japan) and circular lash (Nikon Corporation, Japan) with 300 dpi (dots per inch).From each study model, two lateral photographs were obtained, one on the right side and one on the let, with the buccal surfaces of posterior teeth parallel to each other.All photographs were obtained with the same distance between the object and the lens (31.4 cm), to avoid magniication.
The digital images were inserted into a computer and analyzed with Dolphin Imaging software version 10 (Dolphin Imaging and Management Solutions, Chatsworth, CA, USA).This program magnifies each image by means of their size in dpi.For each variable, two points are marked and the distance between them was calculated by the software.The accuracy of measurements was 0.01 mm.For the statistical analysis, the mean right and left sides of each measurement was obtained.The calculated variables are presented in the following topics.

Molar relationship
The molar relationship was measured from the tip of the mesiobuccal cusp of the maxillary first molar to the mesiobuccal groove of the mandibular first molar (Fig 1).The average of right and left sides was used.

Second premolar relationship
The relationship of the second premolars was measured from the tip of the buccal cusp of the maxillary second premolar to the distal anatomical contact point between the mandibular second premolar and the mesial of the mandibular first molar (Fig 2).The average of right and left sides was used.

First premolar relationship
The relationship of first premolars was measured from the tip of the buccal cusp of the maxillary first premolar to the distal anatomical contact point of the mandibular first premolar and mesial of the mandibular second premolar (Fig 3).The average of right and left sides was used.

Canine relationship
The canine relationship was measured from the cusp tip of the maxillary canine to the anatomical contact point between the distal mandibular canine and the mesial of the mandibular first premolar (Fig 4).The average of right and left sides was used.

Method error
The intra-examiner error was evaluated by taking new measurements of the initial, final and post-retention study models of 15 patients randomly selected, performing a total of 45 pairs of models.The first and second measurements were performed with a time interval of one month.The formula proposed by Dahlberg 11 (Se 2 =Σ d 2 /2n) was applied to estimate the magnitude of casual errors, while systematic errors were analyzed by applying the paired t test, according to Houston. 19

Statistical treatment
Descriptive statistics were performed for the variables at the initial (T 1 ), final (T 2 ) and post-retention (T 3 ) phases, and for the differences between initial and final, (characterizing treatment correction), and tial complete Class II molar relationship.Therefore, the independent t test was applied for all variables between these two groups.
All tests were performed with the STATISTICA software (Statistica for Windows, Release 6.0,Copyright StatSoft, Inc. 2001), adopting a significance level of 5%.

RESULTS
Table 1 presents the results of the evaluation of systematic and casual errors, by evaluating the paired t test and Dahlberg' formula, 11 applied to all studied variables.
The descriptive statistics results (mean, standard deviation, minimum, maximum and number) of the variables molar relationship (MR), second premolars relationship (2PMR), first premolars relationship (1PMR) and canine relationship (CR) in every evaluated stage (T 1 , T 2 and T 3 ) and periods (T 2 -T 1 and T 3 -T 2 ) are shown in Tables 2 to 5.
between the final and post-retention stages, (characterizing the change during post-retention).There was also descriptive statistics of the ages at initial, final and post-retention stages and the duration of treatment, retention and post-retention evaluation.
To evaluate variable changes between phases, the dependent ANOVA for repeated measures (repeated measures ANOVA) was used and in the presence of a signiicant result, Tukey's test was applied subsequently.
The Pearson correlation coefficient was calculated to verify the presence of correlation between molar relationship relapse with: severity of initial Class II relationship, treatment time, retention time and time of post-retention evaluation.
To better evaluate the influence of the initial Class II molar relationship severity on the stability of molar relationship, the sample was divided into two groups: Group 1 with ½ -cusp or ¾-cusp initial molar Class II relationship, and Group 2, with ini-    The choice of the methodology to be used should be based on the objective of the study.Since the purpose of this study was to evaluate the stability and relapse of molar relationship in the post-retention stage, the best method for evaluating molar relationship are study models.Although this method does not allow the clinical and radiographic analysis, the study models alone provide information related to diagnosis and orthodontic treatment. 6,15Furthermore, it has been shown that there is a poor association among the occlusal characteristics and the morphology obtained in the lateral cephalograms and better prediction of orthodontic results can be obtained by occlusal indexes than by cephalometrics. 1 The fact that occlusal characteristics do not always reflect craniofacial morphology was evidenced by Pancherz, Zieber and Hoyer, 35 who observed similar cephalometric characteristics when comparing Class II, division 1 and 2, contradicting the widespread idea that the severe overbite of Class II division 2 malocclusion is related to a more horizontal skeletal pattern of this malocclusion. 21Although this Table 6 shows the results of the analysis of variance (ANOVA) and Tukey's test for the variables molar relationship, relationship of first and second premolars and canines, at the initial, final and postretention stages.
Table 7 presents the results of the Pearson correlation test to evaluate the correlation of the severity of Class II relationship with the post-retention relapse.
Table 8 shows the test results of the Pearson correlation test to determine the correlation of relapse and treatment time, retention time and time of postretention evaluation.
Table 9 presents the results of the independent t test between groups 1 (½-cusp or ¾-cusp Class II) and 2 (complete Class II).

Methods
From the medical records of each patient general data such as date of birth, gender, type of appliance and mechanics used, date of beginning and end of treatment, duration of the use of the retainers, date of post-retention evaluation were collected.With these data, treatment, retention and posttreatment evalua-   Andrews 4 defined the six keys to normal occlusion based exclusively on the information contained in 120 study models, and these six keys are valuable parameters to obtain an ideal static occlusion.Similarly, the occlusal evaluation is an important research tool regarding the results of orthodontic treatments. 2,7,43herefore, this study performed the measurement of molar relationship on study models, a method that is simple and objective, and has previously been used in several studies. 10,12,23,39ll measurements were made from photographs of the study models of the three phases for each patient.The study models were photographed with a D-80 camera, with a 105 mm close-up lens and circular flash (Nikon Corporation, Japan) with 300 dpi.The use of this lens prevents any distortion of the image.From each model, two lateral photographs were obtained, one of the right side and one of the left, with the buccal surfaces of the posterior teeth parallel to each other.The digital images were inserted into a computer and analyzed with Dolphin Imaging software version 10 (Dolphin Imaging and Management Solutions, Chatsworth, CA, USA).This program magnifies each image by means of the size in dpi, to be informed by the examiner.For each variable, two points are marked and the distance between them is thus calculated by the software.The accuracy of measurements was 0.01 mm.For the statistical analysis, the average of the right and left sides of each of the measures was obtained.
The main advantage of this measurement method is that the images, once inserted in the Dolphin software, can be magnified on the computer screen, or even displayed with a multimedia projector, and the points marked with the aid of a mouse connected to the computer.This possibility of image magnification greatly facilitates the visualization of the point to be marked, minimizing the possible methodological errors, as shown in Table 1.
According to Houston, 19 for an accurate analysis, the object of study should be reevaluated a minimum of 25 times.Thus, to evaluate the intra-examiner error, new measurements of the studied variables (molar relationship, relationship of second premolars, re-lationship of first premolars and canine relationship, Figures 1, 2, 3 and 4, respectively) were performed on study models of 15 patients randomly selected from the total sample, a total of 45 pairs of models, measured one month after the first measurement.The results of the two measurements were then subjected to the formula proposed by Dahlberg,11 to obtain the casual errors.To obtain the systematic errors, the paired t test was applied.
The results demonstrated the absence of systematic errors.Casual errors were minimum and hence acceptable (Table 1).The greatest casual error occurred for the canine relationship (CR), with value of 0.27 mm.The absence of significant systematic errors and the minimum value of the casual errors observed in this study may result from both the standardization and accuracy of the measurements, and also by the simplicity and objectivity of the measurement used, making this method very reliable and reproducible.

Sample
Since the main objective of this study was to evaluate the stability of molar relationship in the longterm, the selection of the sample was performed aiming to eliminate the largest possible number of factors that could influence the results.Therefore, to evaluate the stability of molar relationship, and also the relationships of first and second premolars and canines, it was necessary to standardize the initial characteristics and the several factors related to the orthodontic treatment.Therefore, the initial malocclusion was calibrated, regarding the type and minimum severity, the treatment protocol used, and the type of appliance.
Therefore, the basic criteria for sample selection was initially Angle Class II malocclusion, with molar relationship of at least half-cusp Class II. 42Cases could not present rotation of the maxillary and mandibular molars in the initial models, which could influence the measurement of the variables.In addition, all patients should have been treated with fixed orthodontic appliances 14,37 in both maxillary and mandibular arches, without extractions. 14,20,37All patients used headgear in the maxillary arch and Class II elastics during orthodontic treatment.
Cases treated previously with functional, fixed and removable appliances, and with intraoral distalizers were excluded, which may influence the interpretation of the results of this study.It is known that the relapse of the skeletal changes of functional orthopedic appliances occurs after removal of the appliances, and this could influence the results. 32,34In addition, distalizers also perform a quick distalization of the maxillary molars and generally cause a distal tipping of the crown of these teeth, relapse may be increased in these cases, due to these factors. 27,30he presence of permanent teeth erupted up to the first molars and the absence of supernumerary teeth and agenesis constituted criteria of sample selection, since the absence of permanent teeth, the presence of supernumerary and some anomalies related to the shape of the teeth can interfere with the normal development of the occlusion, producing malocclusions that require correction with a different orthodontic mechanics, increasing the complexity and difficulty of the orthodontic treatment, and stability. 6,25he sample consisted of a total of 41 patients, selected from the records of the Discipline of Orthodontics, treated by Graduate students (from the Department of Pediatric Dentistry, Orthodontics and Public Health, Bauru Dental School).Only those cases that had complete orthodontic records were selected, with all the forms properly completed, presenting the study models from initial, final and at least two years posttreatment stages.
The time of posttreatment evaluation, in the postretention stage, is reasonable to observe the stability, one of the purposes of this study, because, according to Al Yami, Kuijpers-Jagtman and van't Hof, 3 about half of the total relapse occurs in the first two years after the end of treatment, with good stability for most of the characteristics in the period of more than five years posttreatment.
The sample selection did not involve the factor quality of finishing, which did not serve as a criterion for exclusion or inclusion.However, assessing the Class II cases treated without extractions in FOB-USP, Barros 6 found that even those cases where there is need for greater patient cooperation, 20,41 were finished, in general, in an acceptable manner.Furthermore, it has been previously demonstrated that the quality of finishing is not related to the long-term results of orthodontic treatments, and an excellent finishing does not guarantee stability. 13,29,33

Molar relationship
The measurement of the initial Class II molar relationship showed a mean value of 4.65 mm, and was reduced to 0.50 mm after treatment.For the postretention evaluation, in the long-term, 0.62 mm was found (Table 2).This demonstrates a correction with treatment of 4.14 mm and a minimum relapse of only 0.12 mm (Table 2).As shown in Table 6, after performing the ANOVA and Tukey tests, a statistically significant correction with treatment and stability in the post-retention period could be noted, since there was no statistically significant difference in the molar relationship between the final and post-retention stages.In other words, the molar relationship showed to be stable in the post-retention phase.
The results of this study are in agreement with previous findings in the literature. 10,12,23,39anut and Arias, 10 evaluating Class II division 2 cases, found a mean of post-retention relapse of molar relationship of 0.6 mm, and all patients had a good molar occlusion in the post-retention phase.The authors considered the molar relationship stable at the end of the post-retention period.
Kim and Little 23 found even an improvement in molar relationship in the post-retention period evaluating Class II division 2 cases.At the end of treatment, the cases had a mean value of 1.3 mm for the molar relationship and, in post-retention stage, this value reduced to 1.2 mm, suggesting an improvement in the molar relationship of 0.1 mm.
Uhde, Sadowsky and BeGole, 39 evaluating Class I and Class II cases in the post-retention stage, reported that the mean change in molar relationship is always in relation to the Class II, however, these changes are not relevant, (about 0.50 mm).However, besides including Class I and Class II cases, which directly influences the results, the authors also included cases treated with and without extractions.
However, Fidler et al 12 found a significant relapse of molar relationship between the final and post-retention stages.However, although statistically significant, this relapse had low values of 0.34 mm for the molar relationship on the right side and 0.33 mm on the left side.Perhaps this difference with the present study was due to the fact that the authors selected cases with Class II malocclusion treated successfully Stability of molar relationship after non-extraction Class II malocclusion treatment original article at the end of treatment, and in the present study, the final treatment outcome of the Class II was not considered for the sample selection.
The literature shows stability of the molar relationship, especially in Class II, division 2 cases.Regarding the Class II division 1 malocclusion, this study showed slightly better results than those found in the literature. 12

Second premolar relationship
The initial measurement of the second premolars relationship presented a mean value of 5.89 mm, being reduced to 1.39 mm after treatment.For the post-retention evaluation, in the long-term, 1.60 mm was found (Table 3).This demonstrates a correction of 4.50 mm, and a relapse of 0.21 mm (Table 3).As shown in Table 6, after performing the ANOVA and Tukey tests, a statistically significant correction with treatment and stability in the post-retention period could be noted, since there was no statistically significant difference of the second premolars relationship between the final and post-retention stages.
The study of Kim and Little 23 showed the same trend, however, in the inal stage, the value of the premolars relationship was slightly higher than normal, and this value remained higher in the posttreatment stage.The initial value of the Class II premolars relationship was 4.6 mm, being corrected to 2.2 mm at the end of treatment and relapsed to 2.5 mm in the post-retention stage.

First premolar relationship
The initial measurement of the first premolars relationship presented a mean value of 5.88 mm, being reduced to 1.03 mm after treatment.For the postretention evaluation, in the long-term, 1.33 mm was found (Table 4).This demonstrates a correction of 4.85 mm, and a relapse of 0.30 mm (Table 4).As shown in Table 6, after performing the ANOVA and Tukey tests, a statistically significant correction with treatment and stability in the post-retention period could be noted, since there was no statistically significant difference in the first premolars relationship between the final and post-retention stages.

Canine relationship
The initial measurement of the canine relationship presented a mean value of 7.22 mm, being re-duced to 2.32 mm after treatment.For the postretention evaluation, in the long-term, 2.28 mm was found (Table 5).This shows a correction of 4.89 mm, and an improvement in the post-retention period of 0.04 mm (Table 5).As shown in Table 6, after performing the ANOVA and Tukey tests, a statistically significant correction with treatment and a complete stability in the post-retention period could be noted, since there was no statistically significant difference of the canine relationship between the final and post-retention stages, and even a small improvement could be seen.
Kim and Little 23 found results similar to the present study.The initial Class II canine relationship had value of 5.3 mm, and at the end it was corrected to 1.6 mm and remained stable, showing the same value of 1.6 mm in the post-retention stage.

Correlations
To verify the correlation between the severity of the Class II relationship with the post-retention relapse, and the relapse of Class II molar relationship with treatment time, retention time and time of postretention evaluation, the Pearson correlation test was used (Tables 7 and 8).
There was a correlation of the initial canine relationship with relapse (Table 7).
There was correlation of molar relationship relapse with the time of post-retention evaluation (Table 8).However, this correlation was negative, indicating that the longer the time of post-retention evaluation, the lower the relapse of molar relationship.This finding seems unreasonable; however, since patients were mostly young, at the end of orthodontic treatment, they still showed growth in the post-retention stage.As the growth tends to improve the relationship of skeletal bases, 18 it is natural that the if time has passed until the post-retention evaluation, the patient will have more growth, favoring the stability of the correction of the Class II molar relationship.

Intergroups comparison
To check the influence of the severity of the Class II molar relationship in the initial stability of the molar relationship, the sample was divided into two groups: Group 1 with initial half-cusp or ¾-cusp Class II molar relationship, and Group 2, with ini-original article Lima DV, Freitas KMS, Freitas MR, Janson G, Henriques JFC, Pinzan A tial complete Class II molar relationship.Therefore the independent t test was applied for all variables between these two groups.
There was compatibility between the two groups for the initial and final ages, and only the post-retention age showed a statistically significant difference (Table 9).Subjects in Group 2 (complete Class II) had an older age in the post-retention stage than the subjects in Group 1.However, despite the complete Class II group had presented an older age in the postretention stage, the time of post-retention evaluation of this group was not statistically significant higher than the group with ½-cusp and ¾-cusp Class II.The retention time was also compatible between the two groups studied (Table 9).
Regarding treatment time, the group with complete Class II showed longer treatment than the ½-cusp and ¾-cusp Class II group, and this difference was statistically significant (Table 9).This was expected since it is known that the severity of malocclusion, especially the severity of the Class II, when treated without extractions, can significantly increase treatment time. 6,20,40egarding the initial molar relationship, as group selection was based on the severity of this relationship, a significant difference between the groups, was expected (Table 9).Obviously, the complete Class II group had a significantly higher value than the group with less severity.The molar relationship at the end of treatment and at the post-retention stage did not differ between the two groups (Table 9).There was also a difference in the amount of correction with treatment, which also was expected, because if the Class II molar relationship was more severe in Group 2, a greater correction of this relationship was really necessary in this group (Table 9).The molar relationship relapse between the two groups did not present a statistically significant difference, however, it was observed that the molar relationship presented a relapse of 0.23 mm in the group with less severity and only 0.04 mm in the group with greater severity.This evidence reinforces the findings of this study that there is no relationship of the initial Class II severity with molar relationship relapse.
The same pattern of results was observed for the first and second premolars and canine relationship.There were statistically significant differences be-tween the two groups for these relationships in the initial (T 1 ) and final stage (T 2 ) and the change with treatment (T 2 -T 1 ) (Table 9).For the relationships at the beginning of treatment, as explained above for the molar relationship, this result was expected due to the greater Class II severity in Group 2. Thus, it was also expected that the correction with treatment was greater, as confirmed by the results.However, there was also a significant result for these relationships at the end of treatment.Group 2, with complete Class II, presented a deficient finishing for the premolars and canine relationships.
A very important factor to be considered is the need for patient cooperation.The successful Class II malocclusion treatment without extractions is extremely connected to patient cooperation.It should be noted that a severe initial malocclusion, demands greater need for patient cooperation to achieve a satisfactory final result. 6,42According to Barros,6 the treatment of a complete Class II without extraction, in relation to two maxillary premolar extractions, requires approximately twice the degree of patient compliance.
Due to the fact that Group 2 includes only complete Class II malocclusion cases, it could be assumed that the greater severity of molar relationship at the beginning of treatment and hence the greatest amount of correction during treatment would influence the maintenance of long-term results. 28However, this association was not confirmed by the present study, since both groups had a stability without statistically significant differences for the molar, premolar and canine relationships.
Araki 5 also found no major posttreatment changes in the group where the change of molar relationship was higher during treatment, and speculated that it probably occurred due to the proper retention of the dental relations obtained.
By studying the stability in the long-term posttreatment of Class II malocclusion by means of the Herbst appliance, Hansen, Pancherz and Hägg 16 observed that, when the maxilla and the mandible are well connected in a stable Class I relationship, the strength of the maxillary growth can be transmitted to the mandible and vice versa.For this reason, it is essential to finish treatment with the best possible intercuspation.

FINAL CONSIDERATIONS
It is considered that the stability of the correction of dental relationships, such as molar and canine relationships, are primary goals of a successful orthodontic treatment.
The stability of dental relationships is the most important aim to be achieved, 36 because relapse was clearly considered, in a clinical assessment of dentists and patients, as the cause of dissatisfaction in relation to orthodontic treatment.The posttreatment changes of the skeletal characteristics have secondary importance, since they are not visible in the clinical evaluation, but should also be aimed, because these changes may reflect changes in tooth position. 5he present study demonstrated a relative occlusal stability of the molar, second and first premolars and canine relationships, as the posttreatment changes were minimal.This is a very important data for planning and treating the cases, because, apart from the cephalometric data of the patients, it is known that at least the occlusal relationship, considered as the most important, is almost maintained in the long-term.
It is also important to highlight that there is a wide individual variability of stability and relapse, since it has a multifactorial cause, and there are several factors related to it, such as craniofacial growth, patient compliance with the use of retainers. 26

CONCLUSIONS
According to the studied sample and to the methodology used, it can be concluded that: » There was a non significant relapse of molar relationship, on average 0.12 mm.The relapses of premolars and canine relationships were also not significant.» There was a significant correlation only between molar relationship relapse and time of post-retention evaluation.» When the sample was divided into two groups, with half-cusp and ¾-cusp Class II and with complete Class II at the beginning of treatment, no difference in the relapses of molar, premolars and canine relationships could be found .» The Class II molar relationship correction remained stable and the initial severity did not influence molar relationship relapse.

Table 1
-Results of t test and Dahlberg formula, applied to the variables molar relationship, second premolar relationship, irst premolar relationship and canine relationship, for estimation of systematic and casual errors, respectively (n=45).

Table 4 -
Results of the descriptive statistical analysis for the variable irst premolars relationship (1PMR), in all stages and periods evaluated.

Table 2 -
Results of the descriptive statistical analysis for the variable molar relationship (MR), in all stages and periods evaluated.

Table 3 -
Results of the descriptive statistical analysis for the variable second premolar relationship (2PMR), in all stages and periods evaluated.

Table 5 -
Results of the descriptive statistical analysis for the variable canine relationship (CR), in all stages and periods evaluated.

Table 6 -
Results of the analysis of variance (ANOVA) and Tukey tests for the variables molar relationship, irst and second premolars relationship and canine relationship (n=39), between the initial, inal and post-retention stages (same letters mean no statistically signiicant diference).

Table 7 -
Results of the Pearson's correlation test to verify the correlation of the severity of the Class II relationship with the post-retention relapse.

Table 8 -
Results of the Pearson's correlation test to verify the correlation of the relapse with treatment time, retention time and time of post-retention evaluation.

Table 9 -
Results of independent t test, between the groups divided in ½-cusp or ¾-cusp Class II and complete Class II.