Ethical approval for this study was obtained by the Universiti Teknologi MARA (UiTM) Research Ethics Committee on September 29, 2017 (Reference: 600-IRMI [5/1/6]; REC/297/17). This trial was also registered retrospectively at the International Standard Randomised Controlled Trial Number (ISRCTN) registry with the study identification (ID) ISRCTN15080404. Participation was voluntary, and withdrawal from the trial was allowed without jeopardizing the patient’s orthodontic treatment.

This study was a prospective, randomized, single-center, two-arm parallel clinical trial to compare the effects of MOPs on EARR during the initial alignment stage of moderate maxillary crowding in comparison with conventional orthodontic treatment alone. The trial was conducted at the Orthodontic Clinic, Faculty of Dentistry, UiTM.

Inclusion criteria included age ranging from 18 to 45 years, moderate crowding of 5–8 mm in the maxillary arch, full dentition up to the second molars prior to extraction of the upper first premolars with or without anchorage reinforcement, healthy periodontal status with probing depth less than or equal to 3 mm, full-mouth plaque score of 20% or lower, and full-mouth bleeding score of 20% or lower. Exclusion criteria included previous orthodontic treatment, systemic diseases, smoking, pregnancy, and compromised periodontal oral health with evidence of bone loss and root resorption on DPT, as well as history of dental trauma. Patients with dental (e.g., hyperdontia, hypodontia) or craniofacial anomalies (e.g., cleft lip and palate) were also excluded. Patients receiving medications that could interfere with tooth movement, such as anti-inflammatory drugs, systemic corticosteroids, or calcium channel blockers were not taken into consideration. The period of recruitment was from October 2017 until February 2019.

Standardized conventional pre-adjusted edgewise McLaughlin, Bennett, and Trevisi brackets with a 0.022 × 0.028-inch (in) slot (Victory Series; 3M Unitek, Monrovia, CA, USA) were used for all participants. The archwire sequence was similar for both groups, which involved the use of 0.014-in nickel titanium (NiTi) archwires followed by 0.018-in NiTi archwires (TruFlex NiTi archwires; Ortho Technology, Lutz, FL, USA) until alignment was achieved.

A single operator (A.A.S.) performed MOPs under local anesthesia infiltration (Lignocaine with 1:100,000 adrenaline). The patients in the MOP group received two perforations 3 mm apart within the attached gingiva to 1 mm apical to the mucogingival junction, with the MOPs placed equidistant between the anterior teeth from the upper right canine to the upper left canine, except in the midline alveolar bone. The perforations were 1.5-mm wide and 3-mm deep and were made using a disposable MOP device (PROPEL Ortho Singapore, Kallang, Singapore). All participants underwent follow-up assessments at four-week intervals, and MOPs were repeated at every visit until completion of the alignment stage. Postoperatively, local measures with cotton roll pressure were applied to arrest bleeding after MOPs. Participants in the MOP group were prescribed 0.12% chlorhexidine digluconate (Oradex; CAVICO, Selangor, Malaysia) mouthwash to be used twice a day for one week. In addition, proper oral hygiene maintenance was reinforced, and the participants were advised to avoid modifying their flossing and brushing routines in the MOP region.

Long-cone intraoral periapical (IOPA) radiographs were used to assess the occurrence of EARR. All images were taken by a standardized paralleling technique using a single digital dental intraoral machine (Satelec X-Mind-AC/DC; Acteon, VA, Italy). IOPA radiographs involving six anterior maxillary teeth were taken before bracket placement, on initiation of MOPs (T1), and after six months into treatment (T2) (Figure 1). Crown length and root length were measured using a virtual ruler of the EasyDent V4 viewer software (Vatech, Hwaseong, Korea) by identifying these landmarks in all three radiographs at each stage (Figure 2). The crown length was gauged from the mid-incisal point of the crown to the median cementoenamel junction (CEJ) point, which was the midpoint between the mesial and distal CEJ. The root length was calculated from the median CEJ point to the tipping point of the root. Apical root resorption was measured as the difference between the length of the root at T1 and the length of the root at T2 in millimeters.

Image magnification between pre- and post-observational radiographs was rectified by exploiting the crown length as a reference, assuming crown lengths to be unaffected throughout the trial interval. Consequently, a correction factor (CF) was formulated by dividing the initial crown length (C1) by the post-treatment crown length (C2). The CF was computed to relate the pre-treatment (R1) and post-treatment (R2) root length of each tooth, as shown in the following equations. Then, the EARR per tooth was calculated in millimeters using the formula adapted from previous studies^16,17:

The grading of apical root resorption was assessed separately by the principal investigator (A.A.S.) in the IOPA images by adapting the scoring system proposed by Levander and Malmgren.¹⁸

The sample size was calculated to detect a significant mean difference of 0.54 mm of root resorption with a standard deviation of 0.47 mm during the alignment stage of fixed appliance treatment after six months.¹⁹ The statistical power was set at 80%, with a confidence interval (CI) of 95%. A minimum of 12 participants for each group was required, and this number was increased to 15 per group considering a 20% dropout rate.

Once the participant agreed to the study protocol, written informed consent was obtained. Participants were allocated randomly to the MOP or control group in a 1:1 allocation ratio using block randomization of six numbers (three odd and three even numbers). Odd numbers were allocated to the MOP group while even numbers were assigned to the control group. Sequentially numbered opaque, sealed envelopes were used for concealed randomization. Cards were written with numbers and set in opaque sealed envelopes, which were kept by the central trial coordinator (N.H.N.), who was responsible for assigning the envelope numbers upon operator request (A.A.S.). The subjects’ participation protocol is shown in the Consolidated Standards of Reporting Trials flow diagram (Figure 3).

Blinding of participants and the operator was not feasible due to the nature of the intervention. Thus, the outcome examiner was blinded to the subject’s assigned intervention. Patient information was concealed by the central trial coordinator (N.H.N.) prior to the measurement assessments.

Descriptive and analytical statistics were obtained using the SPSS software version 25.0 (IBM Corp., Armonk, NY, USA). The statistical significance level was prespecified at p < 0.05, with a 95% CI. Significant changes in EARR were analyzed at the patient level using repeated-measures analysis of variance to compare the mean differences in EARR among groups. Stata software version 13 (StataCorp, College Station, TX, USA) was used for kappa analysis.

Calibration for root resorption was performed on 15 IOPA radiographs to ensure the reliability of the measurement technique. The intra-rater reliability was excellent with an intra-class correlation coefficient of 0.82, with the 95% CI (0.65, 0.91), indicating a low level of random error. Further analysis with a paired t-test revealed no significant difference between the two readings (p = 0.242). Grading for the extent of EARR was also performed with an interval of two weeks on 15 IOPA radiographs, and the intra-rater reliability of Cohen’s kappa was 0.880 with 93.33% agreement (p < 0.001).

A total of 35 participants were eligible for this trial. However, five participants were excluded, including three patients who declined to participate and two who had moved away. The remaining 30 participants who met the inclusion criteria were enrolled for the trial. Subsequently, one pregnant participant dropped out from the MOP group, and one participant in the control group was lost with no follow-up records. An equal number of participants in both groups was maintained throughout the observation period. Thus, analysis as per protocol was undertaken to compare root resorption between the MOP and control groups. The mean age in the MOP and control groups was 22.80 ± 3.78 years and 22.50 ± 2.74 years, respectively, with no significant difference between the groups (p = 0.810) (Table 1).

A comparison of the changes in root length (Table 2) showed no significant difference between the MOP and control groups (p = 0.238 and p = 0.225, respectively). Further analysis between groups also did not show a significant difference (p = 0.419).

Approximately 50% of the teeth in both groups showed no root resorption in the first six months. Most of the roots in the MOP and control groups (42.86% and 52.38%, respectively) showed only mild resorption. Less than 8% of the roots in both groups (7.14% in the MOP group and 4.76% in the control group) showed moderate resorption (Table 3).

Mild resorption is defined as some alteration of the root apex, while moderate resorption indicates less than 2 mm of root resorption.¹⁸ In an extensive study for developing an index for root resorption with an observation period of six to nine months after orthodontic treatment, Levander and Malmgren¹⁸ reported that 66% of the teeth showed zero or minor resorption, 33% showed moderate resorption, 17% showed resorption, and 1% showed extreme resorption. Similarly, in their study of 302 patients, Smale et al.²⁰ assessed root resorption for 4 incisors over 6 months after initial orthodontic alignment and reported that 47.1% of the patients showed no root resorption, 29.9% and 19.4% showed mild and moderate resorption, respectively, and only 3.6% showed severe root resorption. Overall, almost half (47.62%) of the measured teeth showed mild root resorption and only 5.95% showed moderate root resorption. The remaining teeth showed no incidence of severe or extreme resorption.

The MOP and the control groups showed no significant differences in mean root resorption (p = 0.419), and the changes in root length over the six-month period within groups were not significant respectively (p = 0.238 MOP and p = 0.225 in control groups). The present study is the first to compare the EARR on the anterior maxillary teeth during initial alignment with that after MOP application. Our results are consistent with those reported by Alkebsi et al.²¹ for a three-month trial of MOP-assisted orthodontic canine retraction. They found that the difference in mean root resorption between the intervention and control groups, with EARRs of 0.61 mm and 0.73 mm, respectively, was not significant. Similarly, the split-mouth study by Aboalnaga et al.,²² also found no significant difference in EARR in the four-month canine retraction period. The latter study used adjusted cone-beam computed tomography (CBCT) for measurement of root resorption. In contrast, a microcomputed tomography (micro-CT) RCT on the extracted first maxillary premolars following 28 days of buccal tipping with a force of 150 g reported that MOPs significantly increased EARR by 0.17 mm.²³ However, the results of that study should be considered with caution, since the clinical relevance of that technique may be debatable in general clinical practice.

The lesser extent of EARR exhibited by the MOP group is in agreement with the findings of a previous trial.⁹ This is due to regional acceleratory phenomenon activity induced by the disruption of the alveolar bone by MOPs, which causes intensification of osteoclastic activity and reduced bone density. Such a reaction subsequently reduces the possibility of hyalinization necrosis and resorption of the tooth root. Additionally, the increased cortical bone porosity simultaneously reduces mechanical resistance for OTM and EARR.⁹

Most EARR diagnoses are based on radiographic evaluation, with apical blunting or shortening of the root length appearing as the first noticeable signs. Root resorption is measured by using IOPA radiographs with a CF adapted from previous studies to compensate for image magnification.^17,24 One of the limitations of using IOPA is the tendency to underestimate apical root resorption. This is justified by a study performed by Dudic et al.²⁵ that aimed to validate the accuracy of IOPA against a micro-CT scanner method in the diagnosis of EARR on extracted teeth. IOPA detected only 55% of EARRs in the experimental teeth whereas the micro-CT scans detected 86% of the EARRs on similar teeth. The most reliable and valid tool to assess root resorption is CBCT. It allows the roots to be visualized and evaluated on any surface and eradicates the image superimposition. It also ensures accuracy in the measurement and ensures better comparability. Nevertheless, CBCT exposes patients with a higher dose of radiation and is more expensive.²⁶ For these reasons, CBCT endorsement in orthodontics is limited to specific indications.

The effectiveness of detecting EARR by using two-dimensional intraoral radiographs is arguable. Root resorption is a three-dimensional process that can occur at any of the root surfaces. Our EARR evaluation used two-dimensional long-cone periapical radiographs, which could limit the accuracy of root resorption calculation and its occurrence. Even though this approach shows limitations in evaluating all surfaces and tends to underestimate root resorption, the clinical relevance of the undetected EARR on plain radiographic images is debatable. The severity of such EARR is usually clinically insignificant. Moreover, both CBCT and IOPA have been shown to detect a comparable percentage of moderate and severe EARR. Thus, the protocol using IOPA radiographs in this study reflected contemporary orthodontic practice, yielding a meaningful trial outcome pertinent to routine clinical practice.

In general, mild to moderate EARR rarely shows clinical significance unless it is accompanied by severe or extreme resorption. However, root resorption of more than quarter of its length may adversely affect root longevity and yield an unfavorable crown-root ratio.²⁷ Remarkably, Levander and Malmgren²⁸ suggested that long-term tooth survival was not compromised even in cases with major EARR. Sondeijker et al.²⁹ developed a comprehensive clinical practice guideline that recommended different management approaches for EARR before, during, and after orthodontic treatment. Even though the guidelines were established principally from a universal agreement between clinicians with limited existing evidence, they permitted clinicians to act in response to EARR based on present expertise. The guidelines include a strong recommendation to inform patients about the risk of EARR prior to orthodontic treatment, with greater caution to be exercised in extraction cases, which are associated with a greater possibility of severe EARR. They also strongly recommend reassessment and revision of the treatment aims and plan when EARR equal to or more than 2 mm is apparent during fixed appliance treatment.²⁹

No root perforation or soft tissue scarring were identified throughout the trial period in all participants. No accidental adverse effects were reported.

Root resorption is a three-dimensional process that can occur at any of the root surfaces. Our EARR evaluations were performed using two-dimensional IOPA, which could limit the accurate calculation of resorption, since its occurrence is three-dimensional. Future studies should investigate the effect of MOPs throughout the treatment duration to assess its complete effectiveness.

The clinical effectiveness of MOPs on the rate of tooth movement over a broader range of orthodontic treatments remains to be investigated. The alignment duration and pain experienced by the patients with the additional MOPs are described in another part of the study.

The generalizability of this study may be limited by the single-center nature of the clinical trial. However, these findings are representative of general clinical practice, where standard mechanotherapy is performed in conventional orthodontic treatment for adults.