Long-term stability of maxillary anterior alignment in non-extraction cases

Methods: The sample comprised 23 patients (13 female; 10 male) at a mean initial age of 13.36 years (SD = 1.81 years), treated with fixed appliances. Dental cast measurements were obtained at three different time points (T1 – pretreatment, T2 – posttreatment and T3 – long-term posttreatment). Variables assessed in maxillary arch were Little Irregularity Index, intercanine, interpremolar and intermolar widths, arch length and perimeter. The statistical analysis was performed by one-way ANOVA and Tukey tests when necessary. Pearson’ correlation coefficients were used to investigate possible associations between the variables evaluated.


INTRODuCTION
The primary purpose of orthodontic treatments is malocclusion correction; however, treatment stability shows considerable variability during post-retention phase.Despite the literature consensus that some occlusal changes will inevitably occur ater orthodontic treatment, 15,19,28 it is noted that long-term stability of the aligned teeth is highly variable and unpredictable. 17reater research emphasis has been placed on relapse of mandibular anterior crowding and little emphasis has been given to investigating the maxillary crowding relapse and parameters that may be helpful in predicting its long-term stability. 2,3,9,12,18,23,25lignment stability of mandibular incisors is less than that of the maxillary anterior teeth. 8,10,22,26,29,30actors such as pretreatment crowding severity 25 and gingival fibers traction 5,6,7 are considered risk factors for maxillary incisors crowding relapse.However, there is an association between a prolonged period of retention and greater stability of maxillary teeth alignment. 23Maxillary incisors tend to rotate in the direction of their initial positions, 25,26 despite buccolingual relapse being unpredictable. 25Furthermore, palatal contacts between maxillary and mandibular incisors preclude lingual movement of the maxillary teeth and any vestibular movement is probably determined by the lips position and function. 12ccordingly to Little, 14 evidence of progressive instability is often first noted by progressive crowding of mandibular incisors following removal of retaining devices.Whatever the multiplicity of causes for relapse, mandibular incisor irregularity is often the precursor of maxillary crowding, deepening of the overbite, and generalized deterioration of orthodontic treated cases.
Kahl-Nieke, Fischbach and Schwarze, 13 evaluated pretreatment, posttreatment, and post-retention models of 226 cases with all types of anomaly.Findings indicated that relapse of incisors crowding occurred in approximately half of the sample and that post-retention crowding increased more frequently in mandible than in the maxilla.Pretreatment variables such as severe crowding and incisors irregularity, arch length deficiency, arch constriction and increased overbite were found to be associated factors in the process of post-retention increase of crowding and incisors irregularity.Premolars extraction treat-ment exhibited greater maxillary and mandibular crowding relapse than non-extraction protocol.
In a longitudinal study, Moussa, O'Reilly and Close, 18 evaluated 55 non-extraction orthodontic patients that were previously submitted to rapid palatal expansion (RPE).The authors 18 observed that maxillary incisors irregularity increased 0.60 mm during post-retention.They suggested that RPE procedure may be helpful in long-term stability; however, due to the absence of a control group there was no clear evidence about a possible influence of RPE procedure on the crowding relapse.However, Canuto et al, 3 compared the long-term stability of maxillary incisor alignment in patients treated with and without rapid maxillary expansion.They concluded that RME did not influence long-term maxillary anterior alignment stability.
Vaden, Harris and Gardner, 30 concluded that most (96%) of the maxillary incisor irregularity correction was maintained after 15 years of treatment.At the post-retention recall, the maxillary irregularity index increased only 0.30 mm.Surbeck et al 25 evaluated whether pretreatment misalignment of the maxillary anterior teeth are of significance for post-retention relapse of alignment.The results suggested that anatomic contact point displacement of the maxillary anterior teeth and maxillary incisor rotation relative to the dental arch are significant risk factors for postretention relapse of alignment and that the pattern of rotational displacement relative to the dental arch has a strong tendency to repeat itself.
Taner et al 27 evaluated the effects of fiberotomy in alleviating dental relapse of incisors after orthodontic treatment.The authors described that there was significant increase of irregularity index in the control group, for both maxillary and mandibular anterior segments.Meanwhile, in the group where circumferential supracrestal fiberotomy was performed, no significant increase of the irregularity index was noted.One year later, Huang and Artun, 12 evaluated whether post-retention relapse of maxillary and mandibular incisor alignment were associated.The authors suggested 12 that the occlusal contacts with the mandibular anterior teeth represent lingual boundaries for the maxillary incisor movement, and any labial movement is likely to be determined by the position and function of the lips.In addition, also suggested that the post-retention movement of the mandibular incisors may be influenced by the position of the maxillary incisors and vice versa and indicated that an association between the post-retention misalignment of the incisors in the 2 arches might exist.
Ferris et al 9 investigated the long-term post-retention stability of RPE-lip bumper therapy followed by full fixed appliances.The sample comprised 20 patients at the late mixed dentition that were recalled to obtain post-retention records.The subjects were out of retention for an average of 7.9 years.The majority of treatment increases in maxillary and mandibular arch dimensions were maintained during post-retention phase.Post-retention incisor irregularity increased 0.5 mm in the maxillary arch and 1.1 mm in the mandibular arch.The authors 9 concluded that use of RPE-lip bumper therapy in the late mixed dentition followed by full fixed appliances is an effective form of treatment for patients with up to moderate tooth size-arch length discrepancies.
Erdinc, Nanda and Isiksal, 8 evaluated long-term stability of incisor crowding in orthodontic patients treated with and without premolar extractions.Minimal incisor crowding relapse occurred (0.19 mm and 0.12 mm for extraction and non-extraction groups, respectively).Maxillary incisor irregularity relapse was smaller than mandibular incisor relapse for both groups.Intercanine width expanded during treatment.Incisor positions in both groups tended to return to pretreatment values.Clinically acceptable stability was obtained. 8ecause of insufficient studies on maxillary anterior tooth alignment and parameters that may be helpful in predicting its long-term stability, this study aimed to evaluate the long-term maxillary incisors crowding relapse and possible factors that may influence tooth alignment stability.

Material
The sample was obtained from the files of Bauru Dental School, University of São Paulo, Bauru, São Paulo, Brazil, and consisted of Class I and II malocclusion patients treated orthodontically without extractions.
The criteria for sample selection included the presence of all permanent teeth at treatment beginning (at least first permanent molars) and the absence of shape and/or number dental anomalies.All patients had complete orthodontic records, including study models of the initial phase (T 1 ), end of treatment (T 2 ) and post-retention (T 3 ).None of the subjects underwent rapid maxillary expansion.
Sample comprised 69 dental casts of 23 subjects (13 girls and 10 boys; initial mean age: 13.36 years; SD = 1.81 years) who received full maxillary and mandibular fixed edgewise appliances.These patients underwent orthodontic treatment for a mean period of 2.18 years (SD = 0.93) and were satisfactorily finished at a mean age of 15.54 years (SD = 1.86).The post-retention study models were taken after a mean period of 4.92 years (SD = 1.11).
Regarding initial malocclusion, ten patients had Class I, 8 had quarter-cusp Class II, and 5 had half Class II anteroposterior molar relationships.None of the patients exhibited posterior crossbite at T 1 .
After active treatment, all patients wore a full time Hawley retainer in the maxillary arch for 12 months.A lingual canine-to-canine mandibular bonded retainer was placed and left for a mean period of 3 years.

Dental cast measurements
The T 1 , T 2 and T 3 maxillary dental casts were used.All dental cast measurements were made with a centesimal precision digital caliper (Mitutoyo America, Aurora, Ill, São Paulo, Brazil).
All were linear measurements, in millimeters, described as follows: A) Maxillary incisor irregularity 14

RESuLTS
Dahlberg's formula and Paired t tests showed no significant casual and systematic errors.
The Table 1 exhibits results of one-way analysis of variance (ANOVA) with the post-hoc Tukey test (different letters means a statistically significant difference between variables) that were used to determine whether there was a significant difference between the measured variables during T 1 , T 2 and T 3 .The results showed that the incisors irregularity had significant changes not only during treatment but also at posttreatment.Maxillary crowding relapse occurred in most patients with a mean percentage of 30.64% of the treatment correction.However, no significant differences were detected to the dimensional variables evaluated during the 3 phases, except for the inter-first-premolar width (INTERPB), that exhibited a statistically significant increase from pretreatment (T 1 ) to posttreatment (T 2 ).
Results of the Pearson correlation tests are in the Tables 2 and 3.There was a significant and negative correlation between maxillary incisors crowding relapse and the relapse of the intercanine and interfirst-premolar widths.

DISCuSSION
Although incisors alignment relapse in maxillary arch is less prevalent than in mandibular arch, the evaluation of possible factors that may influence maxillary tooth alignment stability has validity.Relapse of crowding in this region may also results in esthetic and functional occlusal deficiencies.Mainly due to its

Statistical analysis Method error
Within a month interval from the first measurement, ten dental casts from T 1 , T 2 and T 3 phases were randomly selected and remeasured.The casual error was calculated according to Dahlberg's formula (Se 2 = Σd 2 /2n). 4The systematic error was calculated with dependent t tests, according to Houston. 11

Statistical method
One-way dependent ANOVA and Tukey tests were used to evaluate the behavior of the measured variables during the three phases (Initial -T 1 ; Posttreatment -T 2 ; Post-retention -T 3 ).
The Pearson correlation coefficient was calculated by using the whole sample to investigate a significant correlation between maxillary incisors crowding relapse and the pretreatment irregularity or the amount of crowding correction.
Pearson correlation coeicient was also calculated to investigate a association between maxillary incisors crowding relapse and the relapse of intercanine, interpremolar or intermolar widths, arch length and perimeter.
The results were considered statistically significant at p < 0.05.All statistical analyses were performed with the software Statistica for Windows, version 6.0, Statsoft, Tulsa, Okla, USA.    1) showed that occurred statistically signiicant changes in the Little irregularity index during the three phases studied.It was observed a signiicant maxillary crowding reduction during treatment.However, there was a signiicant relapse of the incisors irregularity ater treatment.Regarding changes in maxillary arch dimensions during treatment, there was only a signiicant change in the variable INTERPB (Inter-irst-premolar width), suggesting that most maxillary arch dimensions were maintained during treatment, and remained stable during post-retention.Sadowsky et al 23 evaluating stability in maxillary and mandibular dental arches of patients treated without extractions and Edgewise mechanics, observed no signiicant changes in the intercanine and inter-premolars widths, ive years post-retention.Erdinc, Nanda and Isiksal, 8 evaluated stability of the orthodontic treatment, with and without extractions.Similarly to the present study, it was observed signiicant decreases in maxillary incisors irregularity during treatment.Patients treated without extractions exhibited signiicant increases of the intercanine and inter-premolars widths during treatment.The maxillary arch dimensional measurements showed no signiicant changes ater treatment, however, relapse of maxillary crowding was signiicant.These studies 8,23 and the present suggests a favorable prognosis regarding maxillary arch long-term dimensional stability of orthodontic cases treated without premolars extractions.
In the present study, mean post-retention crowding relapse was 1.52 mm.Sadowsky et al, 23 assessing stability of subjects treated non-extraction, reported a relatively similar amount of relapse (1.1 mm), 5 years post-retention.However, Moussa, O'Reilly and Close 18 observed more favorable results regarding crowding relapse, 8 to 10 years post-retention, in a sample comprising 18 subjects treated with rapid maxillary expansion and ixed appliances.It was observed a mean maxillary crowding relapse of 0.6 mm (SD = 1.30).Vaden, Harris and Gardner, 30 noted that 96% of the maxillary crowding correction remained original article Canuto LFG, Freitas MR, Freitas KMS, Cançado RH, neves LS stable 15 years ater treatment.The amount of crowding increased from 1.5 mm at posttreatment to 1.8 mm at post-retention.Ferris et al 9 also evaluated the maxillary crowding relapse in non-extraction cases.It was observed only 0.47 (SD = 1.19) of maxillary irregularity increase during the post-retention (7.9 years).The increased stability observed in these studies may be explained by the prolonged retention protocol ater orthodontic treatment. 23In the Sadowsky et al 23 study, retainers were placed and let for a mean period of 8.4 years.Moussa, O'Reilly and Close, 18 described a mean period of 6.6 years of retention for the mandibular arch and full time Hawley retainer in the maxillary arch for 2 years.The study conducted by Vaden, Harris and Gardner, 30 reports that patients used these retainers in the maxillary and mandibular arches or these retainers in the maxillary arch and a lingual canine-to-canine mandibular bonded retainer.The irst posttreatment control was carried out only ater six years.The study by Ferris et al 9 described a retention protocol that included the use of Hawley in the maxillary for 3 years (full time during one year) and lingual bonded retainer or Hawley plates in the mandibular arch for a mean period of 3 years.In the present study, all subjects wore a full time Hawley retainer in the maxillary arch for 12 months.A lingual canine-to-canine mandibular bonded retainer was placed and let for a mean period of 3 years.
Erdinc, Nanda and Isiksal 8 described an increase in maxillary incisors irregularity of 0.19 mm and 0.12 mm for patients treated with or without extractions, respectively, 4 years and 11 months ater treatment.The extraction group showed 4.4 mm of pretreatment crowding.However, the non-extraction group exhibited only 1.94 mm of initial irregularity.The initial crowding was signiicantly less severe than that observed in our sample (6.56 mm).Maxillary and mandibular retainers were removed at least two years before the post-retention measurements.The exceptional stability of this study may be related to the amount of initial crowding and due to a short interval between retainer removal and the postretention evaluation.
Although the results indicate a posttreatment maxillary crowding relapse greater than that reported in previous studies, 8,9,18,23,30 the mean irregularity index at posttreatment (3.12 mm) is considered clinically acceptable according to Little. 14sults of the Pearson correlation tests showed no significant correlations to most variables evaluated (Tables 2 and 3).It was observed that the amount of initial crowding had no effect on relapse, as described in previous studies. 1,17Surbeck et al, 25 in contrast, reported a positive correlation between the amount of maxillary incisors irregularity and the amount of incisors crowding relapse.The authors reported 25 that the tendency to maxillary crowding relapse increases 2.3 times for each 0.2 mm of incisors contact point displacement in relation to the dental arch.Furthermore, each 4° of tooth rotation at pretreatment has increased by 2.7 times the probability of irregularity relapse.The authors 25 also pointed out that partially aligned tooth exhibits significant risk of relapse.They suggested the use of individualized retention protocols and that patients should be aware about the possibility of relapse accordingly to the initial irregularity. 25However, a positive correlation between the amount maxillary incisors crowding at pretreatment and the crowding relapse after treatment seems unlikely when analyzing our results and previous studies.For example, ours results indicated that the experimental group exhibited 6.56 mm of initial irregularity and had a mean post-retention relapse of 1.52 mm.The mean crowding relapse observed in the present study was higher than that reported by Ferris et al, 9 Sadowsky et al 23 and Vaden, Harris and Gardner, 30 with samples that exhibited more maxillary incisors irregularity at pretreatment (10.45 mm, 8.0 mm and 7.9 mm, respectively).Despite more initial crowding, these studies reported less posttreatment irregularity relapse (0.47 mm, 1.1 mm, 0.3 mm, respectively).
The amount of maxillary crowding relapse (LIT-TLE3-2) showed a statistically signiicant and negative correlation (p < 0.05) with the post-retention changes in intercanine (INTERC3-2) and inter-irst-premolars (INTERPB3 -2) widths (Table 3).These results suggest that the higher the post-retention decreases of intercanine and inter-irst-premolars widths the higher the maxillary crowding relapse.However, although these correlations have statistical signiicance, the coeicients values observed implicate in a weak correlation (r values of -0.459 and -0.419, respectively).Therefore, it can be argued that the observed correlation between relapse of the maxillary crowding and the reduction of these dimensional measurements has poor clinical signiicance.
Long-term stability of maxillary anterior alignment in nonextraction cases original article Moreover, it seems obvious that the reduction of these measurements tends to be consequence of maxillary arch constriction in the anterior region.Therefore, it is expected a space availability decrease and an increase in the amount of tooth crowding.
Despite numerous studies that evaluated a possible relationship between changes in intercanine width and mandibular incisors crowding relapse, the correlation between maxillary crowding relapse and maxillary intercanine width changes were only investigated by Surbeck et al, 25 and Erdinc, Nanda and Isiksal. 8Surbeck et al 25 found a signiicant association (p < 0.001) between intercanine width decreases and maxillary incisors crowding relapse, however, the correlation test result also revealed a weak association (r < 0.70).Erdinc, Nanda and Isiksal, 8 found no correlation between post-retention changes in incisors irregularity and changes in intercanine width.

Clinical implications
Maxillary anterior alignment shows better prognosis regarding stability when compared to the same region in the mandibular arch, this fact may explain the scarce studies in literature about this issue.Despite the greater stability, maxillary crowding relapse can compromise orthodontic results after retention appliances removal.The post-retention relapse observed in this study (1.52 mm), although statistically significant, may be considered clinically acceptable. 14therwise, this minimal amount of crowding relapse can lead to patient dissatisfaction.
Maxillary incisors crowding relapse shows some etiological factors as retention time, initial crowding severity; relapse of teeth in the opposite side, changes in arch dimensions, rotated teeth at pretreatment and lack of complete correction of rotated teeth resulting in absence of adequate interdental contacts.Thus, it becomes clear that more stable results can be obtained with a prolonged retention protocol and an adequate alignment of maxillary incisors during treatment.

CONCLuSION
The maxillary incisors irregularity increased significantly (1.52 mm) five years posttreatment.
None of the clinical factors studied in the dental casts demonstrated to be predictive of the maxillary crowding relapse.
The results suggest that more attention regarding maxillary arch retention protocol should be taken by the clinician.Although alignment stability of mandibular incisors is less than that of the maxillary anterior teeth, maxillary crowding relapse can be signiicant.
(LITTLE) (Fig 1).B) Intercanine width (A; INTERC): The linear distance between the cusp tips of the maxillary canines.When there was a facet, the cusp tip was estimated (Fig 2).C) Inter-premolar widths (INTERPB and IN-TERPB'): The linear distance between left and right central fossae of the maxillary first (B) and second (B') premolars (Fig 2).D) Intermolar width (C; INTERMOL): The linear distance between the mesiobuccal cusps tips of the maxillary first molars.When there was a facet, the cusp tip was estimated (Fig 2).E) Arch length (D + E; LENGTH): The linear distance along the midline from the interincisal midline to the mesial contact of the irst molars (Fig 2).

F
) Arch perimeter (F; PERIM): The distance in millimeters from the mesial dental contact of the left first molars to the mesial dental contact of the right first molars to (Fig 2).

Table 1 -
Results of one-way analysis of variance (ANOVA) with the post-hoc Tukey test (diferent letters means a statistically signiicant diference between variables) for the variables measured on dental casts (N = 23), at the three stages studied (T 1 , T 2 and T 3 ).

Table 2 -
Results of the Pearson correlation test.

Table 3 -
Results of the Pearson correlation test.