Orthodontic root resorption associated with surgical adjunctive procedures for accelerating tooth movement: a systematic review of randomised controlled trials

This systematic review conformed to the Preferred Reporting Items for Systematic Review and Meta-Analyses (PRISMA) statement²⁵ and the Cochrane Handbook for Systematic Reviews of Interventions.²⁶ The review protocol was registered with the PROSPERO international prospective register of systematic reviews, CRD42018111115.

Seven electronic databases were systematically searched and included: PubMed, Medline, Embase, Scopus, CINAHL, and the Cochrane Library. Further searches were performed through Google Scholar, and the Grey literature databases of Worldcat and OpenGrey. The search strategy incorporated MeSH (Medical Subject Headings), Emtree (Embase Subject Headings) and DeCS (Health Sciences Descriptors), keywords, Boolean operators “AND” and “OR”, and truncations, that were adjusted for each database. The search strategy for PubMed may be found in Table I. The full electronic search strategy for the included databases may be viewed online at the PROSPERO review protocol (CRD42018111115).

To ensure that additional studies were not overlooked, hand searching of four selected journals (the American Journal of Orthodontics and Dentofacial Orthopaedics, Angle Orthodontist, European Journal of Orthodontics, and the Journal of Oral and Maxillofacial Surgery) was also undertaken. Ongoing trials were searched from the World Health Organisation (WHO) International Clinical Trials Registry Platform (ICTRP) and Clinicaltrials.gov for relevance. To locate further studies, reference lists of review articles and all included studies were checked. Requests were also sent to experts in the field, to identify unpublished and ongoing studies. Searches were performed with no restrictions on year, publication status, nor language. The results obtained were exported to the reference management software (EndNote v20, Clarivate Analytics, Philadelphia, USA) and duplicates were removed. Searches were initiated in February 2021 and updated in February 2022.

The processes of article selection, screening according to title and abstract, then full-text, data extraction, was undertaken independently by two authors (KS and BK) and reviewed for RoB and quality assessment (KS, BK and HB). For all steps in the review process, instances in which no decision could be reached after discussion between the initial authors, a further author (SC) assisted in the resolution of disputes. A further two authors (DN and CD) were available for guidance and supervision during the manuscript writing process. Titles and abstracts were primarily used to screen the retrieved studies and identify full-text articles relating to external root resorption from surgical orthodontic mechanotherapy. Full-text articles were retrieved and further assessed for suitability and also scanned for publications in which eligibility was uncertain based on the title and abstract alone. Any additional articles were subjected to the pre-specified inclusion and exclusion criteria and assessed for suitability.

Data were extracted from studies that fulfilled the inclusion and exclusion criteria using a specially designed electronic spreadsheet which was piloted and modified prior to use. The following quantitative and qualitative information was extracted from each included study: name of author/s, country of study, year of publication, study design, teeth assessed, sample size, the gender of patients, surgical method, outcome of interest, assessment method (quantifiable measure and instrument) and follow-up duration. Quantitative data presented in tables or text was preferred over graphs or figures to avoid measurement errors from data approximation; however, if data were not presented in tables, the review authors used the specified data extraction protocol to ensure measurement error was mitigated. Data were simplified to include the difference between experimental groups, regardless of the time of follow-up.

The Cochrane Collaboration’s bias assessment tool was used to assess the RoB of included studies.²⁶ The RoB determined the domains selection bias (random sequence generation and allocation concealment), performance bias (blinding of participants and personnel), detection bias (blinding of outcome assessment), attrition bias (incomplete outcome data), reporting bias (selective reporting), and other bias (bias from other sources). Each study was classified as having low (low RoB for all key domains), unclear (unclear RoB in at least one key domain) or high risk (high RoB in at least one key domain). For studies in which an unclear RoB was determined, the authors were contacted to clarify the domain and an appropriate judgement was made. In instances in which the authors did not respond, nor supply sufficient evidence to make a judgement, the domain was left as having an unclear risk.

The principal summary measure was the difference in the amount of OIERR following the SAPATM during orthodontics. The secondary outcome measure included overall treatment time, rates of tooth movement and periodontal-related outcomes such as mobility, increase in probing depth or changes to bony support. To assess the effect of interventions, the association between the surgical variable and root resorption was compared. The Grading of Recommendations, Assessment, Development and Evaluations (GRADE)²⁷ approach was applied to evaluate the quality of evidence for each outcome. The level of certainty was downgraded by one level for serious concerns (or by two levels for very serious concerns) regarding RoB, inconsistency, indirectness, imprecision or publication bias. This was used to determine the level of certainty in the evidence and strength of recommendations as high, moderate, low or very low quality.

Cohen’s kappa statistic²⁸ was applied to determine the inter-rater agreement of the review authors for the screening and inclusion of primary articles and the evaluation of the quality of evidence (KS and BK). Any disagreements were resolved by discussion and consensus of additional authors (HB and SC).

The electronic and manual searches of the literature revealed 1118 citations (including two articles retrieved from citation searches). Following the removal of duplicates, 682 articles were screened by title and abstract and resulted in the acquisition of 99 full-text articles (k = 0.89). Of the full-text articles evaluated against the pre-specified inclusion and exclusion criteria, a total of 13 articles^29–41 fulfilled the requirements for inclusion in the present systematic review. The inter-reviewer agreement of 0.94 for the inclusion of primary articles indicated almost perfect agreement. Figure 1 shows the PRISMA flowchart of the literature selection process.

Figure 1.

The included studies^29–41 were published between 2012 and 2022, with sample sizes ranging from 20 to 52 patients (aged 15–45). All studies^29–41 used either a traditional (surgical bur) corticotomy, piezocision-assisted corticotomy, corticision (using a surgical blade and hammer), or MOPs to accelerate tooth movement. Several studies^{29,30,32,33,37,38,41} utilised a split-mouth design, in which the participant was subject to the intervention and control on separate sides of their mouth. Importantly, there was variance in the diagnostic modality for determining the presence of OIERR: ten of the included studies^{29–31,33–37,39,41} utilised three-dimensional computed tomography (CT), whereas three studies^32,38,40 used two-dimensional intraoral periapical radiographs (IOPA).

The mean patient follow-up period across studies was approximately 26 weeks; however, this varied considerably between studies (12–77.9 weeks).^29–41 A difference between the follow-up period for intervention and control groups was also noted in two studies.^34,40 Table II provides a summary of the characteristics of the included studies, describing the participants, treatment goal, intervention, comparison, outcomes measured and study design (PICOS), sample size and follow-up period.

Data pooling of the included studies was not suitable due to methodological and clinical heterogeneity. As the included studies used different methodologies, outcome measurements and reporting strategies, a narrative synthesis was conducted rather than a meta-analysis.

An overall low RoB was recorded for the majority of included studies.^{29–33,35–39,41} Two studies^34,40 were considered a high risk overall due to a lack of clarity regarding the method of randomisation. The remaining papers^{29–33,35–39,41} utilised an adequate random allocation procedure. All studies had a low RoB for domain missing outcome data, as they fully reported data on their outcomes. All studies had low concerns regarding their RoB for measurement of the outcome. All studies had a low RoB for the domain selection of the reported result. The inter-reviewer agreement value for RoB was 1.0, indicating perfect agreement between the reviewers. The summary of the RoB assessment for included studies is presented in Figure 2.

Figure 2.

Of the 13 studies^29–41 included in the present review (Table III), six studies reviewed parallel groups^{31,34–36,39,40} and seven were split-mouth studies.^{29,30,32,33,37,38,41} These studies comprised 2128 teeth from 240 patients, with samples ranging between 18 and 32 participants. Of the studies, two performed a traditional corticotomy^29,40, 6 produced a corticotomy using piezocision^29,33–37, one corticotomy was created by corticision³⁹ and there were 6 MOP interventions.^{30–33,38,41}

Two of the included studies demonstrated reduced OIERR in relation to SAPATM. Abbas et al.²⁹ assessed 40 canine teeth from 20 patients during canine retraction and found a reduction in OIERR associated with the use of either piezocision or traditional corticotomy when compared to a control side. Whereas Shoreibah et al.⁴⁰ assessed the mandibular incisors of 20 patients and only studied corticotomy during comprehensive treatment. The study reported that patients treated with conventional orthodontics had significantly higher rates of OIERR than those who underwent SAPATM. Interestingly, Alqadasi et al.³³ noted no significant difference in OIERR when MOP was compared with piezocision. A third study showed OIERR both increased on the mesiobuccal root and decreased on the distobuccal roots of maxillary molars during retraction.³¹

OIERR during canine retraction was assessed by 6 studies and comprised 141 patients, with sample sizes ranging from 18 to 33 subjects. Of these studies, Abbas et al.²⁹ assessed corticotomy and piezocision, whereas Alqadasi et al.³³ compared piezocision and MOPs. Raj et al.³⁷ assessed piezocision, and the remaining three studies^30,32,41 compared MOPs and conventional orthodontic treatment. All studies used bilateral closed nickel–titanium springs (exerting 150 g of force) to assist in the retraction of canines.

A 2021 report by Hatrom et al.³⁶ assessed OIERR induced during en-masse retraction of maxillary anterior teeth by comparing piezocision and conventional orthodontic treatment. This parallel design study³⁶ placed 12 patients into the SAPATM group, and 11 patients in the control group. The results showed that, while statistically significant OIERR was detected in both groups, there was no statistically significant decrease in OIERR following SAPATM compared to controls.

The study by Alkasaby et al.³¹ assessed OIERR during the retraction of maxillary first molars when comparing SAPATM of flapless MOPs with conventional orthodontic treatment. This parallel design study placed 12 patients into the SAPATM group, and 11 patients in the control group. OIERR was significantly greater in the mesiobuccal roots but significantly decreased in the distobuccal roots of the first molars that underwent SAPATM. The palatal root changes were insignificant between experimental and control groups.

Orthodontic levelling and alignment of the dentition were assessed by five studies, and consisted of 132 patients.^{34,37,38–40} These studies used SAPATM techniques of piezocision,^34,35 MOPs,³⁸ corticision,³⁹ and corticotomy⁴⁰ techniques in conjunction with orthodontic treatment in both arches. Three studies^34,35,39 used CT as their assessment method for OIERR, whereas two studies^38,40 utilised IOPA. Only Shoreibah et al.⁴⁰ noted a statistically significant reduction in OIERR following the use of SAPATM during orthodontic levelling and alignment.

Secondary outcome measures including a statistically significant decrease in overall treatment time was shown in four of the included studies,^34,35,39,40 with an additional six papers showing an improved rate of orthodontic tooth movement in the SAPATM groups compared to controls.^{29–31,33,37,41} A study by Abbas et al.²⁹ showed that a corticotomy produced greater rates of canine movement than piezocision during canine retraction. Of the deleterious effects of SAPATM, transient pain^30,32 and scarring^34,35 was reported in some papers. Despite these negative effects, two studies from the same author found statistically significant increases in patient satisfaction with SAPATM.^34,35 Statistically significant changes to the periodontium during SAPATM noted a decreased bone density in the middle and apical thirds of root length³¹ decreased canine palatal bone height following piezocision³³ and a decrease in the relative attachment level.³⁷

The overall certainty of the evidence for each intervention was judged as ‘low’ to ‘very low’ using the GRADE tool. This grading suggested very little confidence in the effect estimate for the reviewed outcomes. Therefore, the true effect was likely to be substantially different from the estimated effect for most interventions. Table IV outlines the reasons for these judgements. The inter-investigator agreement (kappa) value for the assessment of the quality of evidence was 1.00, indicating perfect agreement between reviewers.

While it is widely accepted^4,14,16,24 that the use of orthodontic force often results in unwanted OIERR on the treated dentition, there are limited high quality, well-designed studies measuring the effect of SAPATM on OIERR. Adjunctive corticotomy use in orthodontics is regularly studied, although most publications are case presentations and are used to highlight procedural effectiveness in accelerating tooth movement and safety profile.⁴² SAPATM may be introduced clinically due to its reported decrease in overall treatment time, by 1.5–4 times in some studies.^43,44 These findings are supported by the present study, with four of the included studies,^34,35,39,40 showing a statistically significant decrease, with an additional six papers describing an improved rate of orthodontic tooth movement in the SAPATM groups compared to controls.^{29–31,33,37,41} Though most included studies^{29,30,32,33,37,38,41} used a split-mouth design, effective randomisation allowed for external validity of the statistically significant decrease in treatment time found in these papers. Additionally, Abbas et al.²⁹ showed that a corticotomy produced greater rates of canine movement than piezocision during canine retraction.

The primary outcome of the present study was to evaluate the effect of SAPATM procedures on OIERR. SAPATM has been shown to be associated with a reduced risk of OIERR and other deleterious dental, or periodontal sequelae.^16,42 Although often associated with greater forces than conventional second phase orthodontics, SAPATM has been histologically shown to result in less OIERR and associated hyalinisation.⁴⁵ This suggests that SAPATM may reduce or eliminate cortical resistance, or increase local metabolism of the periodontium, resulting in decreased periodontal ligament pressure.⁴² Despite this, only two of the 10 included studies showed a statistically significant reduction in overall OIERR when SAPATM was employed during orthodontic tooth movement.^29,40 Interestingly, Alkasaby et al.³¹ found that first molars which underwent SAPATM had a statistically significant increase OIERR in the mesiobuccal roots, whereas the distobuccal roots displayed a significant reduction. A positive correlation between bone density and root resorption was noted by the study, and it is postulated that this, in combination with the shorter distance of the distobuccal root from the MOP site, contributed to the significant reduction.

Current evidence suggests that osteoclast activation during orthodontic tooth movement is related to the intracellular levels of monocyte chemotactic protein-1, chemokine ligand 3, 5 and interleukin-8.^46–48 Importantly, this rise in inflammatory mediators following SAPATM has been shown to only persist up to 28 days following the surgical procedure.⁴⁹ This suggests that the benefits displayed by SAPATM may be considered temporary, with the duration of conventional orthodontic treatment offsetting any initial gain. This potentially explains the findings of the present review, as the experimental duration of the studies by Abbas et al.²⁹ and Shoreibah et al.⁴⁰ were 12 and 17.5 weeks, respectively.

A similar process to the regional acceleratory phenomenon (RAP) may also explain the limited number of studies which show a statistically significant reduction in OIERR during SAPATM. While the mechanism behind RAP appears to be due to an increase in cytokines, which are responsible for the recruitment and activation of osteoclasts, some evidence suggests that these inflammatory cytokines may circulate throughout the body and result in a ‘systemic’ acceleratory phenomenon.^50–52 Therefore, the reduction in root resorption exhibited in the included studies may be explained by a ‘systemic’ acceleratory phenomenon on the control side originating from the RAP associated with SAPATM on the experimental side.⁵³ Particularly with the split-mouth design studies,^{29,30,32,33,37,38,41} the ‘systemic’ acceleratory phenomenon may help explain why the majority of studies showed no statistically significant reduction in OIERR when SAPATM was performed.

All included studies measured the change in root length; however, two different measurement tools were applied by ten studies^{29–31,33–37,39,41} which used CT and three which used IOPA.^32,38,40 It may be argued that CBCT is more accurate in the standardisation of the measurements throughout treatment, in addition to comparing each side on the same patient. From the studies that used IOPA,^32,38,40 it is difficult to determine the accuracy and reproducibility of the employed technique by the clinician responsible for reporting the radiographs. In relation to CBCT analysis, multiple software programs were used across all studies and only one paper³⁵ explicitly stated the CBCT slice thickness (0.2 mm). Due to individual treatment goals, the follow-up period varied greatly for the included studies.^29–41

The present review only included RCTs, as they are considered to be experimentally superior, while alternative study designs potentially provided results which may skew the findings. However, the limitations of the current study include the overall heterogeneity of the included studies, which precluded a meta-analysis.^29–41 Although various force mechanisms, assessment methods and time-points were experimentally used, the present review assessed the overall range of mean difference. During the RoB assessment, two studies^34,40 were downgraded after multiple attempts were made to contact their respective authors for methodological clarification. The present study did not exclude any papers on the basis of RoB, and as no secondary meta-analysis could be performed, the overall impact of the unclear and high RoB studies on the overall treatment effect was not determined.

As the current studies^29–41 performed SAPATM concurrently, it is unclear whether any purported reduced treatment time would account for the time required to plan and perform surgery. In Australian healthcare systems in which orthodontic surgical treatment is commonly performed by oral and maxillofacial surgeons,⁵⁴ the time required to plan and perform the surgery may increase the overall treatment time, and offset any benefit gained.

Several studies^55–58 analysed the effects of low-level laser therapy, photobiomodulation and light-emitting diodes (LEDs) on orthodontic tooth movement. These studies show adjunctive therapies which demonstrated a statistically insignificant decrease in OIERR. Whilst novel, the research team did not classify these techniques as SAPATM and they were therefore excluded.

Despite the use of an exhaustive search strategy, the inclusion of high-quality clinical trials was limited. The GRADE tool showed that the overall certainty of the evidence for each SAPATM intervention was determined to be low (Table IV). It is recommended that further studies, overcoming the methodological limitations of the included trials, are performed.

Although SAPATM has been shown to significantly reduce treatment time, the evidence of its effect on OIERR is limited and would be considered of low quality. In relation to the primary and secondary outcomes of the present study, and based on the best available current evidence, it may be proposed that SAPATM can be undertaken without detrimental effects on the tooth or surrounding structures, whilst also providing several benefits to both clinician and patient, in the form of a decreased treatment time.