Dental extractions performed in orthodontics are followed by space closure, which can be done either with the older frictionless approach (using closing loops that exert light continuous force but are difficult to fabricate and use) or with the friction-based method that uses sliding mechanics, and despite wasting energy and anchorage, it is highly predictable, convenient, and common. ¹ , ² , ³ , ⁴ , ⁵ , ⁶ , ⁷ , ⁸

However, friction is usually undesirable and can compromise treatment and cause complications. ⁹ , ¹⁰ , ¹¹ Friction is suggested to be associated with factors such as the cross-section of arch wire, bracket design, angulation of wire with bracket, materials used in arch wires and brackets, and the type of ligation. ⁹ , ¹⁰ , ¹¹ , ¹² , ¹³ Regarding materials, although stainless steel (SS) brackets have usually higher practical qualities, they are not esthetic; ¹⁴ this is crucial to many patients nowadays, ¹⁵ leading to the introduction of ceramic brackets. ¹¹ , ¹⁵ Albeit these brackets look appealing, they have higher coefficients of friction and cause greater frictions and induce slower tooth movements in comparison with SS brackets. ⁹ , ¹⁰ , ¹¹ , ¹² , ¹⁵ , ¹⁶ , ¹⁷ , ¹⁸ The effect of ligation method (elastomeric/steel ligatures, active/passive self-ligating brackets [SLBs]) on friction ¹⁹ , ²⁰ , ²¹ influences tooth movement rate. ²² , ²³ SLBs are suggested to reduce friction and overall treatment time, improve plaque control, deliver forces in more biological levels, enhance patient comfort, and increase clinical efficacy. ¹³ , ²³ , ²⁴ , ²⁵ , ²⁶ , ²⁷ , ²⁸ , ²⁹ Besides, in comparison to conventional ceramic brackets, metal slot ceramic brackets and self-ligating ceramic brackets might give improved clinical performance, resembling that of conventional metal bracket systems. ¹⁰ However, controversy exists over many of these claims, ⁶ , ²³ , ²⁴ , ²⁶ , ³⁰ , ³¹ and a 2018 systematic review concluded that SLBs might not act better than conventional brackets (CBs) in terms of space closure, expanding transversal dimensions, or orthodontic efficiency. ³¹ Another issue with SLBs that will affect the sliding is binding which is the contact angle of the wire and the wings or the clit of the brackets. ³

Compared to metal SLBs, ceramic SLBs are relatively new to orthodontic practice, and literature on their efficacy is scarce and limited to merely in vitro studies. They may have advantages such as their esthetic look while producing lower frictions. ³² , ³³ This study was conducted since (1) clinical studies on SLBs are scarce and highly controversial, (2) since the few clinical trials are mostly concerned with the speed of retraction and not the other aspects of treatment such as rotation control, (3) because studies on bodily movement of canine during retraction with SLBs are very scarce or nonexistent, (4) since studies on transverse arch changes following canine retraction using SLBs are nonexistent, and finally (5) since there is no clinical study regarding ceramic SLBs. We aimed to comparatively assess the retraction speed, anchorage loss, rotation control, and tipping control of ceramic self-ligating and conventional ceramic brackets for the first time. In addition, for the first time, we used precision three-dimensional (3D) digital measurements for this purpose. The null hypotheses were (1) the lack of any effect of either bracket type on five treatment outcomes, as well as (2) the lack of any significant difference between ceramic CB and ceramic-active SLB types in terms of the amounts of changes over 3 months in the case of five clinical measurements.

Trial design

This is a preliminary, parallel-arm, single-blind split-mouth randomized clinical trial. Protocol ethics were approved before commencement, by both the Ethics Committee of the University (ethical code: IR.AJUMS.REC.1397.254) as well as an International Online Review Body (RCT code: IRCT20180710040414N1) in accordance with the Declaration of Helsinki. Patients could leave the study anytime they wanted to and treatments would be completely delivered anyways.

Participants, eligibility criteria, and setting

Patients were included from the attendees to the Department of Orthodontics, Ahvaz Jundishapur University of Medical Sciences, Ahvaz, Iran, in 2018–2019. The intervention began on September 6, 2018, treatment for all patients finished on November 17, 2018, and all follow-ups finished on March 17, 2019. The inclusion criteria were the indication for extraction of maxillary first premolars followed by canine retraction, alignment of maxillary central incisors in the leveling and alignment stage, absence of any systemic diseases and any history of using medications that could interfere with orthodontic treatment, absence of any periapical lesions and root anomalies before treatment as confirmed by panoramic radiography, and consent of patients or their parents to participate in the study. The exclusion criteria were debonding of brackets from canines or anchorage teeth, removal of retracting coil springs during treatment, and irregular follow-up visits of patients. Patients with asymmetrical arches were excluded from the study.

Preintervention treatment

This step was carried out to control for confounding variables that could affect canine rotation on both sides. The teeth were bonded using MBT slot 22 brackets (G&H, USA), and attachments were placed on the first and second maxillary molars. The first and second maxillary molars and the second maxillary premolar were connected on each side using ligature wires. The leveling and alignment stage continued up to 18” nickel titanium wires (NiTi, Gestenco, Göteborg, Sweden). The order of NiTi wires depended on the patient: some patients needed more steps and some others needed only 18” wires. Afterward, bilateral maxillary first premolars were extracted. To prevent possible rotation and retraction of maxillary central teeth, due to the presence of transseptal fibers during canine retraction, bilateral aligned maxillary central teeth were attached to each other using a steel wire, and a stop was placed between the maxillary centrals. This stop also prevents wire movement within brackets and maintains the similarity of the arch on both sides.

Randomization and blinding

Randomization was done by a third person using an online random number generator: To remove the nonselection error of the ceramic and ceramic self-adhesive bracket in the right or left canine teeth, a simple randomization method was used to extract random numbers by a third person using the website www.randomizer.org. The patients and statistician, but not the operating orthodontist, were blinded to the treatment allocations. The statistician and patients did not have any access to random assignments. The analyst received coded data. Although the SLB and CBs were not exactly the same, the patients were not aware that which bracket shape is SLB or CB.

Interventions

Canine brackets were replaced with the experimental brackets (active ceramic SLB with a 22” metal slot and a ceramic clip [Quicklear, Forestadent, Germany]) and control brackets (metal-lined ceramic bracket [Encore, Ortho-technology, USA]). Brackets of other teeth were not replaced. The control bracket was attached to the wire using an O-ring (clear, YahongOrtho, Beijing, China). NiTi wires remained for a month until the intervention. Before beginning the intervention Figure 1, alginate templates were taken from the maxilla to record the initial condition and rotation of canines. The impression was poured with dental stone. NiTi wires were replaced with 18” SS wire (Ortho-Technology, USA). Canine retraction was performed using a NiTi closed coil spring (G&H, USA) 9 mm in length that exerted a moderate 150-g force. The force was assessed and adjusted each month. After 3 months of canine distalization Figure 2, a second alginate impression was taken. During this period, patients were visited each month to make sure that the appliances were intact, and the coil forces were appropriate. Figure 1

The occlusal view of the maxilla before beginning the retraction of the canine.

Participant flow

About 320 patients were originally screened before reaching the desired sample size. During the study, three patients who had been originally included were dropped out, all due to a failure in an orthodontic appliance in one of the sides (i.e., bracket hook and coil spring); they were replaced by three new patients (with randomized maxilla sides) Figure 7. Figure 7

The flow diagram of the maxillary quadrants (and not the patients) in this study.

The findings of this study indicated that both methods caused canine distalization, anchorage loss, canine rotation, canine tipping, arch expansion at the canine site, and arch constriction at the premolar site during canine retraction. However, compared to conventional ceramic brackets, self-ligating ceramic brackets would cause slower canine retraction, less rotation, less tipping, and less arch expansion at the canine site. Both bracket types may have similar extents of anchorage loss and similar extents of arch constriction at the premolar site. Since there was no study on ceramic SLBs, we are limited to discuss our results in light of few studies available on metal SLBs.

Metal SLBs have been shown to have the lowest frictions in vitro, ¹³ , ³⁸ , ³⁹ , ⁴⁰ , ⁴¹ , ⁴² , ⁴³ which suggests a faster tooth movement, because less extents of force are wasted on friction and more is available to the tooth movement. Nevertheless, clinical evidence was quite controversial in this regard. Sirinivas ⁴⁴ observed faster canine retractions with SLBs compared to CBs. Similarly, Hassan et al. (using passive SLBs) ³⁷ and Jayachandran et al. (using interactive brackets) ⁴⁵ found significantly greater rates of canine retraction in their SLB groups. A reason for their result is suggested to be their methodology of assessing the canine tip, which can be also influenced by tilting, ²³ because tilting can move the canine tip faster while the body of the tooth has not moved as much. On the other hand, da Costa Monini et al. ⁴⁶ who had measured the distances between the canine tips, found no significant difference between metal SLB and CBs in terms of canine retraction. They published another study in 2017 ²⁹ on the same active SLB and again observed no significant differences between conventional and active SLB in terms of canine retraction. ²⁹ Miles (using passive SLB) ⁴⁷ as well found no significant difference in speeds of both methods while retracting the anterior teeth. Furthermore, DiBiase et al. (using passive SLBs) ⁴⁸ and Deguchi et al. ⁴⁹ did not detect a significant difference between SLB and CB in terms of treatment duration and number of visits. Mezomo et al. ²³ found similar rates of retraction as well; they used passive brackets [although they had not stated this in their paper, the type/brand used implied being passive]. According to Wong et al., ⁵⁰ differences in the extent of space closure between CB and passive SLBs over 3 months were trivial and nonsignificant. There was also a study by Burrow ⁵¹ that, similar to our study, found a slower speed of retraction in their passive metal SLB group compared to CBs. The controversy might be attributed to the assessment of canine tip which can be affected by the tooth movement as well as its tipping; therefore, if tipping occurs, a greater range of movement will be detected. Moreover, different brands in use can differ in terms of quality and various physical and mechanical properties. This will lead to different results in various studies.

In addition, if SLB can exert a better control over tooth movement, the movement would be mostly bodily and hence less tipping would be expected, which translates into a shorter distance of canine tip movement within a fixed time frame. Our study showed that ceramic SLBs can induce bodily movement. The only two available studies on tipping of the canine during retraction have shown that passive and active SLBs may induce significant bodily movements. ⁴⁶ , ⁵² Finally, not all in vitro studies agree that friction is less with SLBs. Some in vitro studies indicate that although there might be some significant differences between SLBs and CBs at lower extents of force, the resistance to sliding would be similar for SLBs and CBs at clinically relevant forces. ²⁷ Another study found greater frictions in their passive SLB group compared to CBs. ⁵² Moreover, although SLBs have usually low friction rates, ³⁸ , ³⁹ active SLBs (which were used in this study) may have considerably higher friction rates than passive ones. ⁵³

The lower rates of friction are suggested to also contribute to a greater control over anchorage and reduce the loss of anchorage. ³⁷ , ⁵⁴ Our results in this regard were consistent with those of da Costa Monini et al. ²⁹ who used active SLBs as well as Mezomo et al. ²³ and Miles ⁴⁷ who researched passive SLBs, and none found a significant difference between CB and SLBs in terms of anchorage loss. However, Hassan et al. ³⁷ and Jayachandran et al. ⁴⁵ reported smaller amounts of anchorage loss in their passive and interactive SLB systems, respectively. These differences can again be attributed to various methodological differences such as shape and size of wires. In earlier studies, SS wires of 18” or 0.025” × 0.019” or 0.025” × 0.017” had been used. ¹³ , ³⁸ , ⁵⁵ We used 18” SS wire because larger wires would press the clip of the active SLB and increase the friction; further, this wire size would improve rotation control when engaged within the 22” slot of the SLBs used in this study. ⁵⁵ Another major source of discrepancy can be the limited time of many studies. For instance, in the 3-month course of this study, canines were retracted much more on the conventional side compared to the SLB side. In this situation with SLB canines moving less, the extent of anchorage loss was comparable in both groups. It is not known if we allowed the canine of the SLB side to move as much as the canine did during 3 months in the conventional side; it would be possible for the anchorage to become lost more on the SLB side.

Canine rotation during retraction is a result of the applied orthodontic forces passing not through the center of resistance of the tooth. ²³ Two of the available clinical trials suggest that rotation control may be better with passive SLBs. ²³ , ³⁷ Moreover, Hassan et al. ³⁷ found a significant improvement in rotation control with passive SLBs, similar to our results. They attributed their finding to their larger diameter of arch wires in comparison to the previous two trials. ³⁷ The above-mentioned methodological issues such as the differences in physical properties of appliance parts can as well account for the dispute. Characteristic differences in brackets of both sides such as materials that used for their construction can affect the results of treatment as well.

Effects of canine distalization using SLBs have not been studied before. We found that SLBs can expand the arch at the canine site, but about half of the expansion caused by CBs. Both bracket types similarly constricted the arch at the premolar site. Our findings pertaining to the CB was similar to those of Aksakalli et al. ³⁵ who observed an expansion of the arch after using CBs. However, they found also expansion in the premolar area, which was in contrast to our findings related to CB (or SLB). Still, in their study, the arch expansion at the premolar area was much smaller than that in the canine area, and even some patients had shown arch constriction like in this study. ³⁵ These findings should be re-assessed for other SLB or CB types so that generalization can become possible.

This study was limited by some factors. The 32 quadrants in 16 patients might not look like a large sample; however, it was large enough: the sample size was predetermined to provide powers considerably >90%. The highly significant results and very high post hoc powers obtained in our study confirmed the adequacy of the sample size and the powers for most tests, which were above 90%. Some trials published recently had only 10 patients assessed over 2 months, still with sufficient powers obtained. ³⁵

This trial was powered only for the canine retraction, but there were other variables included. It was not known before commencement of the trial whether the study could be adequately powered for those as well. Still, the post hoc powers clearly show that almost all tests had powers above 90%.

The trial had a surrogate end point of how much space closure is achieved within a time frame. The actual clinical relevant end point would be the duration of space closure or even better the duration of the whole treatment. However, it was not possible to test such parameters in a split-mouth design.

One might argue that the age was not specified in the inclusion criteria and the study included both growing and nongrowing patients and that the rates of tooth movement are different between these groups and they might not be combined or assumed to produce similar results. Nevertheless, this study was a split-mouth trial, meaning that each patient was his or her own control. This intra-individual matching rules out the effect of age on the difference between experimental (SLB) and control (CB) sides. Choosing a broader age range (as done in this study) could also improve the generalizability which is favorable. Yet, the generalizability of results might be limited to the specific physicomechanical characteristics and brands of the materials used in this study as well as characteristics of the sample such as age range and ethnicity of patients. This limitation was something natural and shared by all previous investigations. Another potential concern of readers can be “not using the main SLB/CBs through the whole treatment and instead using a similar bracket for both sides before the beginning of intervention.” Of course, the answer would be that the use of different brackets before beginning of the intervention would simply disrupt the baseline symmetry between the sides. Moreover, we needed the sides to be treated similarly until the intervention. Another argument could be the use of metal ligatures (instead of O-rings) to exert better control over rotation. We used O-rings since, unlike metal ligatures, they were standardized. Moreover, O-rings and metal ligatures might be both ineffective in controlling rotation. ⁵⁵ Furthermore, it was better to adopt more than two types of brackets. However, a limitation of split-mouth designs is that they usually disallow the inclusion of more than two groups. Randomization concealment was not done in this study; however, given that both sides of a patient's mouth are matched, it might affect the results minimally.

Both bracket types succeeded to distalize the canine at the expense of its rotation and anchorage loss. Compared to the used conventional ceramic bracket with a metal slot, this particular brand of active ceramic SLB exerted a better control over canine rotation and its tipping. However, it reduced the speed of canine retraction compared to the CB, perhaps partly because of the bodily movement it caused. The extent of anchorage loss was similar for both groups during the fixed time of 3 months, and noting the lower extent of retraction occurred in the SLB group. The arch was expanded at the canine site, and this expansion was faster for CBs compared to SLBs. The arch was constricted in the premolar area at a rather similar rate for both bracket types.

Financial support and sponsorship

This work was financially supported by grant: U-97061 from Vice-Chancellor for Research Affairs of Ahvaz Jundishapur University of Medical Sciences. This paper is issued from thesis of Dr. Nasim Ghorani.

Conflicts of interest

The authors of this manuscript declare that they have no conflicts of interest, real or perceived, financial or nonfinancial in this article.