Trial Design

This paper was developed according to the CONSORT (Consolidated Standards of Reporting Trials) statements. It consists in a blind secondary data analysis from a previous randomized controlled trial (RCT) with a 1:1 allocation ratio design, explained in the flowchart (Figure 1).¹⁹ It was not necessary to make any changes after trial commencement.

View Figure

• Download as Powerpoint
• Download Figure

Participants, Eligibility Criteria, and Setting

The study was approved by the ethical committee of the Hospital for Rehabilitation of Craniofacial Anomalies (protocol number CAAE: 41327415.2.0000.5441).

One hundred patients with bilateral cleft lip and palate from 8 to 10 years of age were recruited from September 2011 to September 2013 for a previous study.¹⁸ Sixty patients attended the appointment. Written informed consent was obtained from participants' parents or legal guardians before their recruitment.

Eligibility criteria were: (1) middle or late mixed dentition; (2) both sexes; (3) lip repair performed between 3 and 6 months of age and palate repair performed between 12 and 24 months of age; (4) presence of maxillary constriction and need for maxillary expansion previous to the secondary bone graft; (5) permanent first molars without extensive restorations; (6) good periodontal health; (7) no previous orthodontic treatment; and (8) absence of syndromes or other craniofacial anomalies. Exclusion criteria were: (1) early loss of both first and second maxillary deciduous molars on the same side; and (2) previous secondary alveolar bone grafting. Forty-six patients (34 male; 12 female) were selected according to the inclusion and exclusion criteria (Figure 1).

Interventions

Clinical procedures/interventions were performed by two orthodontic residents under the supervision of two experienced orthodontists of the institution during the period from October of 2011 to February of 2014. The RME group included 23 patients (16 male; seven female) with a mean age of 9.25 years (SD = 1.35), who received maxillary expansion with the Haas-type or Hyrax appliance (Figure 2A). Bands were preferentially adapted on maxillary first permanent molars. When second deciduous molars were banded, a lingual extension wire was placed to the partially erupted maxillary first permanent molar. C-shape clasps were bonded on the deciduous canines. The 11-mm screw (Dentaurum, Ispringen, Germany) was activated two turns twice a day (1 mm/d) until overcorrection. The expansion active phase ranged from 7 to 14 days. The appliance was maintained as a retainer for 5 months. The mean amount of expansion was 5.11 mm (SD = 2.74) for the intercanine (3-3) and 3.53 mm (SD = 2.73) for the intermolar (6-6) regions.¹⁸

View Figure

• Download as Powerpoint
• Download Figure

The SME group included 23 patients (18 male; five female) with a mean age of 9.15 years (SD = 0.99) who received the Quad-helix appliance (Figure 2B). Bands were preferentially adapted on the first permanent molars, or on the second deciduous molars when the first permanent molars were partially erupted. The appliance was constructed using 0.036-inch round stainless steel wire. Activations were performed extraorally at installation and every 60 days until overcorrection. The amount of activation for each appointment was 6 mm in the anterior and posterior regions. The active expansion phase ranged from 4 to 21 months with a mean time of 11 months. For both groups, the overcorrection criterion was occlusion between the palatal cusps of the maxillary teeth and the buccal cusps of the mandibular teeth. The appliance remained in the mouth as a retainer during 5 months after the active phase. The mean amount of expansion was 6.58 mm (SD = 3.58) for the intercanine (3-3) and 3.98 mm (SD = 2.41) for the intermolar (6-6) regions.¹⁸

CBCT was performed using the i-CAT machine (Imaging Sciences International, Hartfield, Pa) immediately before expansion (T1) and after the retention period, when the expander was removed (T2). The CBCT exams were taken for a dentoskeletal evaluation previously published.¹⁹ All CBCT scans were taken with 120 kVp, 8 mA, 26.9 s of exposure time, FOV of 13 cm and a 0.25 mm voxel size. For this study, lateral cephalometric images were reformatted using Dolphin software (Chatsworth, Calif). The head position was standardized with the Frankfurt plane parallel to the horizontal plane in a lateral view. In a frontal view, the infraorbital plane was positioned parallel to the horizontal plane. In the axial plane, the ethmoidal septum was positioned parallel to the vertical plane. Cephalometric measurements are shown in Figures 3 and 4. During the analysis, the examiner was blinded.

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

Outcomes

The outcomes evaluated in this study were sagittal and vertical cephalometric changes of RME and SME in children with BCLP.

Sample Size Calculation

To detect a minimum intergroup difference in vertical growth (SN.GoGn) of 2°, with a standard deviation of 1.7°,²⁰ with an alpha error of 5% and a test power of 80%, a sample of 22 patients was required for each group.

Randomization, Allocation Concealment, and Implementation

Computer-generated randomization based on random permute blocks of 20 patients was accomplished using Stata Software (StataCorp, College Station, Tex) to ensure equal distribution of participants in the groups. Allocation concealment was achieved with sequentially sealed, numbered, opaque envelopes containing the expansion modality allocation cards, which were prepared before the trial. One operator was responsible for opening the next envelope in sequence and implementing the randomization process.¹⁸

Blinding

Blinding of patients and operator was not possible regarding the modality of expansion; however, the outcome assessment was blinded because CBCT-derived cephalometric images were unidentified during analysis.

Error Study

The same examiner repeated the measurements in 50% of the sample after a 30-day interval. The examiner did not participate in any of the clinical activities nor in the randomization process. Reliability was calculated using Intraclass Correlation Coefficient (ICC).²¹

Statistical Analyses

Normal distribution of variables was verified using Kolmogorov-Smirnov tests. Intergroup comparison was performed with t tests. The level of significance used for all the tests of the study was 5%, with an associated 95% confidence interval. Statistical analyses were performed with Statistica software, version 11.9 (StatSoft Inc., Tulsa, Okla).

Main Findings in the Context of the Existing Evidence

CBCT-derived cephalograms have been widely studied and used in orthodontics.^22–26 The method using cephalometric reconstruction from CBCT images showed overall landmark identification errors comparable with conventional digital cephalograms and improvement in the accuracy of cephalometric measurements.^22,27 Patients from this study had pre-existing CBCTs taken for a previous RCT, so CBCT-derived cephalograms were used to eliminate the radiation exposure associated with additional conventional radiographs.

Changes in the anteroposterior position of the maxilla may occur after expansion in noncleft patients.^25,28 In patients with UCLP, no anterior maxillary displacement was found when the Hyrax appliance was used.²³ In patients with BCLP, arch length showed a slight decrease after expansion due to premaxillary retrusion.^18,19 In the current study, no differences were observed in the sagittal outcome of the maxilla between RME and SME. Both Haas/Hyrax and quad-helix appliances produced slight retrusion of the maxilla (Table 2).

In this study, the mandible showed slight and similar retrusion in both groups (Table 2). In noncleft patients, similar behavior has been reported after RME and SME.²⁹

There was similar improvement in the skeletal Class II relationship for both groups (Table 2). This behavior was different from noncleft patients, who usually demonstrate a significant increase in skeletal Class II relationship, with a significantly greater increase after RME.²⁹ This was probably caused by a lack of maxillary growth in patients with BCLP.^2–6 The continuous mandibular growth, characteristic of patients in this age range, may have contributed to the increase of the maxillomandibular difference.³⁰

Although RME causes temporary mandibular clockwise rotation and anterior face height increase,^20,28,31,32 no differences in the vertical outcomes were observed between RME and SME (Table 2). In both groups, slight increases in SN-GoGn, FMA, and LAFH were noticed. The similarity between RME and SME can be explained by the expansion overcorrection of 2 to 3 mm that produced a cusp to cusp occlusion in both groups, increasing the mandibular plane angulation and anterior face height.

The maxillary incisors underwent slight palatal tipping in the RME group and slight labial tipping in the SME, however, without significant intergroup differences (Table 2). In noncleft patients, there is no unanimity regarding maxillary incisor response after RME or SME.^{28,29,31–34} The mandibular incisors remained practically unchanged after expansion in both groups. These findings were in agreement with other studies in noncleft patients.^32,34,35 This is probably because the forces of these expander devices were applied only to the maxillary teeth.

Changes in the soft tissue were similar in both groups (Table 2). The nasolabial angle slightly increased, probably due to the slight maxillary retrusion produced by the appliances, associated with the progressive retrusion of the maxilla during growth caused by the primary plastic surgeries performed in the first months of life.^3–7

The quad-helix had similar cephalometric changes compared to the Hyrax appliance. The inconveniences of slow expansion include the greater expansion time and a less practical clinical procedure due to the constant need for appliance removal/rebonding for activations.¹⁸ These aspects are very relevant when selecting the appliances in clinical practice.

Limitations

The greater treatment time with the quad-helix compared to the Hyrax expander was a limitation of the study concerning growth comparison. However, the difference in treatment time was only 3.8 months and should not have compromised the results.¹⁹ Further studies with longer follow-up times to evaluate the stability and relapse tendency of the cephalometric results after rapid and slow maxillary expansions are necessary.