Introduction

Posterior crossbite is a nonself-correcting common clinical malocclusion often associated with transverse maxillary deficiency and functional mandibular shift. If left untreated, it can lead to the development of craniofacial asymmetries and mandibular dysfunction [13, 21, 22]. Rapid maxillary expansion (RME) is reported to be an efficient clinical technique aiming to correct maxillary transverse deficiency and crossbite [1, 17].

The Haas appliance is a well-known device designed to expand the palate. It is a tooth- and tissue-borne appliance attached to four teeth and to the palatal vault (Fig. 1). High forces are generated during RME and they can affect the periodontal and endodontic status of the anchoring teeth [24]; therefore, some authors [4, 5, 15, 20, 23] have suggested banding the Haas-type rapid maxillary expander (H-RME) to deciduous teeth. If the roots of the upper second deciduous molars have at least the same length of their crowns at the orthopantomogram diagnostic examination, the H-RME anchored to deciduous teeth is an effective device to correct posterior crossbite [23].

Fig. 1
figure 1

a H-RME banded on upper second deciduous molars. b H-RME banded on upper first permanent molars

Abb. 1 a H-RME befestigt an den oberen zweiten Milchzahnmolaren. b H-RME befestigt an den oberen ersten bleibenden Molaren

Full size image

The aim of the current investigation was to evaluate changes on lateral cephalometric head films and on dental cast models after rapid maxillary expansion with H-RME anchored to deciduous teeth versus H-RME anchored on upper first permanent molars.

Patients and methods

The present paper is a secondary outcome analysis based on a previous multicenter, randomized trial (trial registration: ClinicalTrials.gov Identifier NCT02798822, https://clinicaltrials.gov/ct2/show/NCT02798822). Details on the experimental design, study groups, and treatment methods were previously published [23].

Briefly, a sample of 70 consecutive children (31 boys and 39 girls; mean age 8.4 ± 1.1 years) presenting unilateral posterior crossbite were recruited at the Orthodontic Departments of the Universities of Genova, Siena, and Insubria (Varese), Italy. All the subjects exhibited a Class I or Class II dental malocclusion with ANB <5° and were selected before the pubertal peak (CVM 1–3) [2]. The inclusion and exclusion criteria are shown in Tables 1 and 2.

Tab. 1 Inclusion criteria
Full size table
Tab. 2 Exclusion criteria and number of patients
Full size table

Patients were randomly assigned (using a stratified blocked randomization) to group GrE (H-RME with bands on second deciduous molars; Fig. 1a) or Gr6 (H-RME with bands on first permanent molars; Fig. 1b). After placing of H-RME, the activation rate of the screw was one quarter turn a day (0.22 mm) until overcorrection was achieved, then the appliance was left in situ for 10 months. Pretreatment records (T0) were obtained by means of dental casts, panoramic radiographs, and lateral cephalometric head films. The same set of records was also taken at the removal of H-RME (T1), 10 months after that the device was used for retention purposes. The average treatment time was 12 ± 1.3 months.

Cephalometric analysis

Cephalograms were traced digitally by a single examiner (CC) using Nemoceph 2D software (Arroyomolinos, Madrid, Spain). Each subject was appointed a random identification number so the examiner would be blinded to the subject when measuring. Landmark location and accuracy of the anatomic outlines was verified by a second senior clinician (AU). Cephalometric points selected are shown in Fig. 2 together with the angular measurements used in the present investigation. To analyze the error of the method, five randomly selected lateral cephalometric radiographs were retraced. A combined error of landmark location, tracing, measurement was determined. Intraclass correlation coefficients (ICC) were calculated to compare within-subject variability to between-subject variability. Correlation coefficients for the skeletal measures were greater than 0.94. Linear measurement errors averaged 0.3 mm [standard deviation (SD) 0.6 mm] and angular measurements averaged 0.6° (SD 0.5°).

Fig. 2
figure 2

Angular measurements analyzed. 1 SNA, 2 SNB, 3 ANB, 4 Sn/GoGn, 5 AnsPns/GoGn, 6 Sn/AnsPns, 7 NSAr, 8 SArGo, 9 ArGoGn, 10 ArGoN, 11 NGoGn, 12 U1/Sn, 13 U1/AnsPns, 14 IMPA, 15 U1/L1

Abb. 2 Analysierte Winkelmessungen: 1 SNA, 2 SNB, 3 ANB, 4 Sn/GoGn, 5 AnsPns/GoGn, 6 Sn/AnsPns, 7 NSAr, 8 SArGo, 9 ArGoGn, 10 ArGoN, 11 NGoGn, 12 U1/Sn, 13 U1/AnsPns, 14 IMPA, 15 U1/L1

Full size image

Measurements on three-dimensional dental casts

Maxillary and mandibular casts were processed by means of a three-dimensional (3D) scanner (NextEngine, Inc., Santa Monica, CA, USA) and landmarks considered in this study were traced by means of Rhinoceros 4.0 (Robert McNeel and Associates, Seattle, WA, USA). Measurements were subsequently calculated directly on scanned dental casts by means of an ad hoc software, and first maxillary molars and central and lateral upper incisors rotations were assessed. Each subject was appointed with a random identification number so that the examiner would be blinded to the subject when measuring.

The reference points on the scanned dental casts were the mesiopalatal and the distobuccal cusps of the first upper right and left molars. A line connecting the tips of these two cusps of each molar was used to assess the mesiorotation of each upper permanent molar, as stated by Ricketts [19]. The software traced these lines and also measured the angle formed by their intersections (molar rotation angle, MRA; Fig. 3). The same protocol was also applied to assess the rotation of the right and left central incisor (upper central rotation angle, U1RA) and of the right and left lateral incisors (upper lateral rotation angle, U2RA), taking as a reference the most distal and mesial points of the incisal edge of the central and lateral incisors. The angular measurements formed by the intersections of these lines running on the incisal edges were also calculated.

Fig. 3
figure 3

a Line 1 connects the distovestibular cusp and the mesiopalatal cusp of the upper right and of the upper left first molar; line 2 represents the incisal edge of the upper right and of the upper left central incisor; line 3 represents the incisal edge of the upper right and of the upper left lateral incisor. b Angle A MRA: the inferior angle originated from the intersection of both lines 1; angle B U1RA: the inferior angle originated from the intersection of both lines 2; angle C U2RA: the inferior angle originated from the intersection of both lines 3

Abb. 3 a Linie 1 verbindet den distovestibulären und mesiopalatinalen Höcker des oberen rechten und des oberen linken ersten Molaren; Linie 2 repräsentiert die Schneidezahnkanten des oberen rechten und des oberen linken zentralen Inzisivus; Linie 3 repräsentiert die Schneidezahnkanten des oberen rechten und des oberen linken lateralen Inzisivus. b Winkel A = MRA: unterer Winkel, Ursprung in der Schnittstelle beider Linien 1; Winkel B = U1RA: unterer Winkel, Ursprung in der Schnittstelle beider Linien 2; Winkel C = U2RA: unterer Winkel, Ursprung in der Schnittstelle beider Linien 3

Full size image

In order to assess the angular error of methods, ten maxillary casts were randomly selected and retraced. Intraclass correlation coefficients (ICCs) were calculated to compare within-subject variability to between-subject variability. Correlation coefficients for the angular measures were greater than 0.91 with measurement errors averaged 0.7° (SD 0.6°).

Statistical analysis

Descriptive statistics were computed for all analyzed variables. Shapiro–Wilks test showed that data were normally distributed (W = 0.93). GrE and Gr6 changes at T0 and at T1 were compared using Student’s t tests for independent samples. Probabilities less than 0.05 were accepted as significant in all statistical analyses (p < 0.05). The effect size (ES) coefficient (d) was also calculated [3]. An ES of 0.2–0.3 might be a “small” effect and, thus, have a small clinically significant difference, 0.5 had a “medium” effect, and 0.8 to infinity a “large” effect.

Results

Cephalometric measurements of the two groups reported at T0 and T1 are summarized in Table 3. At T0, none of the skeletal cephalometric variables were significantly different between GrE and Gr6. At T1, no statistically significant skeletal differences were found between the two groups.

Tab. 3 Cephalometric measurements of the two groups reported at T0 and T1 and T test analyses at baseline
Full size table

Between T0 and T1, both groups exhibited a decrease of the upper central incisors angulation on SN line and on the palatal plane; that decrease was found to be statistically significant in GrE when compared to Gr6 (U1/SN −3.1°, p = 0.022; U1/AnsPns −2.8°, p = 0.028). Conversely, lower central incisor angulation on the mandibular plane tended to increase in both groups between T0 and T1, and GrE reported a statistically significant increase of the IMPA when compared to Gr6 (IMPA +2.5, p = 0.029).

The ES result was medium (0.5 < ES > 0.6) for all of the three statistically significant variables (Table 4).

Tab. 4 Unpaired t tests for the net difference T1–T0 regarding the cephalometric variables
Full size table

Dental cast analysis at T0 showed no statistically significant differences between groups regarding MRA, U1RA, and U2RA. Between T0 and T1, MRA decreased in both groups, more in GrE than in Gr6 (−16.1° vs −6.6°). The difference between the groups was statistically significant (p < 0.00) with a very large effect size (ES = 2.4). Moreover, U1RA and U2RA increased in both groups, more in GrE than Gr6 (central 13.4° vs 8.9°, and lateral 18.2° vs 11.5°). The difference between the groups was statistically significant (p = 0.002 for U1RA; p = 0.008 for U2RA) with a ES for incisor changes (ES > 0.75; Table 5).

Tab. 5 Unpaired t tests regarding the net difference T1–T0 for the variables MRA (molar rotation angle), U1RA (upper central rotation angle U1RA), U2RA (upper lateral rotation angle)
Full size table

Discussion

In the present study, after RME treatment, GrE showed a major significant spontaneous retraction (U1/SN −3.1°, p = 0.022; U1/AnsPns −2.8°, p = 0.028) and alignment of the upper incisors (central 13.4° vs 8.9°, p = 0.002 and lateral 18.2° vs 11.5°, p = 0.008) compared to Gr6. This finding is in accordance with Habeeb et al. [9] who reported significant posterior movement of the upper incisors following RME therapy. The effect size found was medium (0.5 < ES > 0.6) for the dental retraction of the upper incisors on the lateral head film evaluation and large (ES > 0.75) for the U1RA and U2RA in the scanned model analysis. These variations are probably due to the significantly greater increase of the intercanine width in GrE versus Gr6 at the end of treatment [23]. As already reported in our previous investigation [23], this effect, in turn, is probably due to the design of the RME in GrE that comprises a more anterior positioned screw.

The major expansion in the anterior area may guarantee for more space for the upper central and lateral incisors that are therefore free to align themselves and mesiorotate spontaneously, also under the influence of the upper lip.

Halazonetis et al. [10] found a three-times increase of buccal pressure on the maxillary first molars after RME. The authors also reported no return to the pretreatment level of pressure regarding the soft tissues during the 3–4 month retention period. Küçükkeleş and Ceylanoğlu [11] showed that the pressure values of the upper lip on the buccal side of upper first molar and incisors increased significantly right after expansion but started decreasing during retention. On the other hand, tongue pressure on the lingual side of the upper first molar and upper incisor decreased significantly with expansion but started increasing after the expansion procedure. This is in accordance with the theory of equilibrium of Proffit [18] that can explain the spontaneous retraction and alignment of the upper incisors. In our study, we can therefore speculate that the lip pressure on the upper incisors could be more present in GrE rather than Gr6 because of the major intercanine width obtained in GrE after expansion and at the end of the retention period. Also Mew [14] and later Mutinelli et al. [16] reported an associated improvement in dental alignment following RME by means of the Little irregularity index measured in the upper arch.

We also have to underline that the values concerning U1RA and U2RA are only analyzed in terms of net difference between the two groups instead of considering the mean value. We considered the mean value less important in this case, considering also that the significant increase of U1RA and U2RA at T2 in GrE was not always represented by a symmetrical movement between the right and the left correspondent incisor. Further analyses are still needed in order to clarify the asymmetrical response of mesiorotation of the right and left upper incisors.

IMPA increased also significantly more between T0 and T1 in GrE compared to Gr6. The retraction of the upper incisors could have reduced an eventual lip interposition between the upper and lower incisors, therefore, creating the lip bumper effect [23]. Nevertheless, the lip interposition was not recorded in our study; therefore, further analysis regarding this parameter needs to be done in order to clarify this event.

In our study, the first maxillary molars distorotated significantly more in GrE than in Gr6 with a very large effect size (ES = 2.4). This could be explained by the triangular opening of the palatal suture due to the position of the center of resistance of the maxilla with respect to the screw position [6, 7, 15, 25]. Moreover, in GrE the upper first molars are free to adapt to the best occlusal situation, since they are not banded.

Our investigation lacks a control group; however, Mutinelli et al. [16] reported that the intercanine width and the intermolar width of the patients who had undergone expansion were significantly higher than the values measured for an untreated control group exhibiting lateral crossbite in the same dental period. Moreover, when compared to the treated group, the control group showed a higher irregularity index in the area of the upper incisors. All these findings indicate that crossbite and upper incisor misalignment do not improve spontaneously with growth in the control groups exhibiting the same characteristics as the treated group before the expansion procedure is performed.

Clinically, the secondary outcome of early treatment of posterior crossbite regarding the spontaneous upper incisors alignment may induce a better arrangement of the transeptal fibers, thus, reducing the probabilities of severely rotated incisors [12]. The spontaneous distorotation of the upper first molars was also detected in our study, in accordance with the literature [15]. Clinically, this result implies a significant increase of the upper arch length [15], a possible improvement of a class II malocclusion [8], and a less invasive and less difficult second phase of treatment.

Conclusions

  • There was an improvement of the anterior crowding and spontaneous retraction of the upper incisor after RME, significantly more in GrE compared to Gr6. This is probably due to more pronounced expansion in the anterior area and a more accentuated pressure of the upper lip in GrE.

  • GrE showed a more significant distorotation of the upper first permanent molars compared to Gr6. This is probably due to the design of the H-RME in GrE, where the screw is more anteriorly positioned and the bands are absent on the upper first permanent molars which are, therefore, free to adapt to the best occlusal situation.

  • Apart from the dental variables measured in the lateral cephalometric head films, GrE and Gr6 did not show any statistical significant difference concerning the skeletal variables.