Effect of pharmacological agents on relapse following orthodontic tooth movement:A systematic review of animal studies

Orthodontic relapse (OR) is an inevitable and unfavorable sequela of orthodontic treatment, which is believed to be multifactorial in etiology. The causes include those beyond the level of teeth and supporting structures, such as neuromuscular imbalance, continued craniofacial growth, and retained habits.^1,2  Some authors emphasized recognizing OR as a distinct manifestation of physiological recovery in response to withdrawal of forces, different from developmental and age-related changes.^3,4 

Over the years, it has been generally accepted that the elasticity of overstretched supra-alveolar connective tissue fibers of the gingiva (primarily the transseptal fibers) exerted physiologic forces causing OR.⁵  This hypothesis was, however, refuted through recent studies, which concluded that periodontal tissue reorganization was rapid and therefore did not contribute to the etiology of relapse.⁴  Further histologic studies, conducted predominantly on animal models, demonstrated that the pattern of tissue and molecular level responses immediately after withdrawal of orthodontic force were similar to those seen during active orthodontic tooth movement (OTM).^6–8  Continued activity of osteoclasts in the direction of relapse, osteoblasts in the direction opposite to that of relapse, gene expression of markers of alveolar bone and collagen remodeling were consistent with that observed during OTM, but in the opposite direction.^6–8 

Underpinned by this similarity in physiological patterns occurring during OTM and OR, chemical preparations determined to decrease the rate of OTM by targeting the cascades involved in alveolar bone remodeling can potentially modulate OR. Evidence from animal and human studies indicated that commonly prescribed drugs, such as nonsteroidal anti-inflammatory drugs, bisphosphonates, and statins, had an inhibitory effect on the rate of OTM.^9,10  In this context, a systematic review of animal evidence on the effect of commonly prescribed pharmacological substances on OR would render an estimate of their effects in humans. This review is the first to aim at appraising the evidence from animal experiments on the effect of pharmacological agents on OR.

The objective of this systematic review was to investigate and evaluate the quality of evidence from animal experiments pertaining to the effect of pharmacological agents on tooth movement relapse following cessation of orthodontic force.

Based on the recommendations outlined in the Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) protocol statement, a protocol was developed and registered with PROSPERO (CRD42019135536).¹¹  The review was conducted in accordance with The Cochrane Handbook for Systematic Reviews of Interventions and reporting conforms to the PRISMA statement.^12,13 

Criteria for determining the eligibility of studies were defined according to participants, intervention, comparison, outcomes, and study design (Table 1). Experiments involving healthy animals subjected to relapse without retention following active tooth movement were assessed. The amount of OR was measured following local or systemic administration of pharmacological agents commonly prescribed to humans and compared with controls with placebo intervention or no intervention.¹⁴  Studies without a control group and reviews were excluded.

Seven online databases, including gray sources, were searched without restrictions on language or date until the third week of April 2019. A detailed search strategy developed for EMBASE (via OVID) by P.V. was later adapted appropriately for other databases (Table 2). Reference lists of all included and excluded articles were searched to ensure that no relevant studies were missed.

The assessment of articles obtained was done by two reviewers (P.V. and S.R.T.) independently and in duplicate. First, the titles and abstracts of articles were assessed for eligibility, following which full-text reports were obtained for articles that were deemed to have met the inclusion criteria.

Relevant data from the included studies were extracted by two reviewers (P.V. and S.R.T.) in previously developed and piloted forms: author and year, study design, subject characteristics, mechanisms inducing OTM, OR characteristics, pharmacological interventions, outcome assessments, and results.

Assessment for risk of bias (RoB) was done by P.V. and V.M. in duplicate using SYRCLE's tool¹⁵  following Makrygiannakis et al.¹⁰  The overall RoB was summarized for each study as specified in the Cochrane Handbook.¹² 

It was planned to use random effects meta-analysis to determine pooled estimates and relative effects of interventions on OR.¹⁶  However, the lack of homogeneity in the study design and interventions employed precluded quantitative data synthesis.¹² 

Although planned, analyses for small-study effects, publication bias, or exploratory subgroup analyses were not feasible because of the inadequate number of studies. The quality of evidence from the retrieved studies was assessed using Grading of Recommendation Assessment, Development, and Evaluation (GRADE) approach, which was graded as high, moderate, low, and very low.¹⁷  The grading represented the confidence in the effect estimate and was based on the following aspects: RoB, inconsistency, indirectness, imprecision, and others.

During the entire process, any disagreements were discussed and resolved by consensus and, when required, a third reviewer (S.G.) was consulted for arbitration.

The process of screening of records is shown in the PRISMA flowchart in Figure 1. The search identified a total of 2354 studies, of which 1633 studies were screened for titles and abstracts after the removal of duplicates. From the remaining 10 records that were assessed for eligibility, 3 studies were excluded because one included a period of retention before experimental relapse,¹⁸  1 involved animals with pathological conditions,¹⁹  and the full-text record of 1 study could not be retrieved.²⁰  Finally, seven studies were included in the qualitative synthesis.^21–27 

Table 3 presents the general characteristics of the studies included. Rats were employed as an animal model in all the studies except one²⁶  that used a rabbit model. OTM was induced using springs placed between either incisors and molars, or incisors and premolars, delivering forces ranging from 50 to 100 cN for a period of 14 to 21 days. One study used elastic bands to effect OTM.²¹  The type of tooth movement achieved was tipping in most of the studies.

The pharmacological agents employed in the included studies were from the following therapeutic categories:¹⁴  antihyperlipidemic agents^{22–24,26,27} ; bisphosphonates²¹ ; analgesics²⁵ ; and antibiotics.²⁷  Administration was systemic in six studies,^21–25,27  whereas local administration was employed in one.²⁶  Systemic administration involved oral, intraperitoneal, or intravenous. Substances were administered either one time, daily, or weekly, for a maximum period of 4 weeks.

All of the included studies measured relapse as the distance moved by the experimental tooth following cessation of active orthodontic force. Timeline of tooth relapse measurements varied from 0.5 days to 4 weeks.

The overview of the assessment of RoB for the included studies is presented in Table 4. All of the studies were assessed as presenting with an unclear RoB and only one study of low risk.²⁶  Most of the studies presented an unclear RoB pertaining to random sequence generation, allocation concealment, blinding of caregivers/investigators, selection of animals for outcome assessment, and blinding of outcome assessor, as there was inadequate information to arrive at a definitive judgment on RoB. However, with respect to the random housing of the animals, one study presented a high RoB.²¹  The majority of the studies included groups that were comparable at baseline for age, gender, and weight and as a result were assessed as having a low RoB. Data from all the animals were analyzed in all of the studies receiving a low RoB rating for the corresponding domain. Finally, it was deemed unlikely that bias could have been introduced as a result of selectively reporting the outcomes and other potential sources.

Table 3 presents the results from the included studies. There was a consistent decrease in the amount of relapse for the experimental period following the administration of pamidronate disodium,²¹  atorvastatin,²⁴  aspirin,²⁵  and chemically modified tetracycline (CMT)-3.²⁷  Inconsistent effects were observed with administration of simvastatin.^22,23,26  The results pertaining to the simvastatin group in the study by Vieira et al.²⁷  was not included in the synthesis considering the significant overlap with a previous study.²³ 

The quality of retrieved evidence on the effects of investigated substances on OR, as appraised using the GRADE approach, was at best low (Table 5).

In summary, based on the evidence retrieved from the eligible animal studies, pharmacological agents commonly prescribed to humans had a variable effect on tooth movement relapse following removal of active orthodontic force. Considering that the available evidence was assessed to have low quality at best, findings from the included studies require careful deliberation. Additional evidence from high-quality studies is required for making definitive recommendations for extrapolation of these findings to human settings.

Theoretically, any substance modulating the signaling pathways involved in periodontal tissue remodeling can affect tooth movement relapse.^4,7,8  Accordingly, it was observed in the current review process that substantial research on the effects of various biomimetics, recombinant proteins and pharmacological substances on OR is being undertaken. However, contemplating the viability of translation of findings to human settings, the current review was restricted to studies investigating commonly prescribed medications.¹⁴  It was, however, surprising to find a scarcity of such research.

OR was shown to decrease with the systemic administration of pamidronate disodium,²¹  atorvastatin,²⁴  aspirin,²⁵  and CMT-3.²⁷  The anti-osteoclast effect of bisphosphonates and consequent inhibition of OTM is well established in the literature.^28–30  Bisphosphonates prevent osteoclastic bone resorption by interfering with cytoskeletal organization and the formation of ruffled borders, inducing apoptosis of osteoclasts, thereby modulating the remodeling of alveolar bone responsible for OR.³¹  Although systemic administration of bisphosphonates was shown to reduce OR, the associated deleterious effects on bone turnover in other tissues must be carefully regarded. Local administration of bisphosphonates was found to enhance the rate of bone regeneration comparable to that of systemic administration with minimal systemic effects.³² 

Statins such as atorvastatin have been shown to influence bone turnover by upregulating bone formation and inhibiting resorption.^33,34  Although lipophilic simvastatin has been suggested to exert greater influence on bone turnover when compared with other statins belonging to same class such as atorvastatin,³⁴  inconsistent effects on OR were noted with the administration of simvastatin.^22,23,26,27  There was a dose-dependent effect of simvastatin on the magnitude of relapse; systemic administration of a low dose of simvastatin decreased OR,²²  whereas systemic administration of a high dose and local administration of a low dose did not affect OR.^23,26,27  Differences in sensitivity of osteoclasts and osteoblasts to different concentrations of simvastatin, experimental animal model, bioavailability associated with route of drug administration, duration of exposure, and experimental designs have been attributed to this varying response of bone remodeling.³⁵  Further studies are warranted to investigate these assumptions.

CMTs have been shown to have beneficial effects in preventing periodontal tissue breakdown, owing to the inhibitory effects on proinflammatory cytokines and matrix metalloproteinase–mediated bone resorption.^36–38  The elevated expression of matrix metalloproteinase in response to stresses in the periodontal ligament tissue during tooth movement alludes to the contributory role of this class of proteinases during relapse.^39–41  Nonsteroidal anti-inflammatory drugs such as aspirin prevent prostaglandin E2–mediated osteoclastogenesis by inhibiting cyclooxygenase activity.⁴²  Aspirin also reduces the release of proinflammatory cytokines such as TNF-α and IFN-γ responsible for promoting bone resorption.⁴³ 

This review was the first to have appraised the evidence on the effects of pharmacological agents on OR. The methodology employed in the current review followed well-established guidelines. An exhaustive search strategy was used for retrieving data from electronic, manual, and gray sources until April 2019. It was comprehensive, without restrictions on date, language, or status of publication. The screening of records for eligibility, data abstraction, RoB assessment, and appraisal of the evidence were conducted in duplicate, and disagreements were discussed or arbitrated with another reviewer to eliminate any possible bias in the methodology.

The limitations of this review arose intrinsically from the study characteristics and retrieved data, resulting in the quality of evidence being assessed, at best, as low. The lack of adequate data precluded quantitative synthesis. Furthermore, most of the included studies were assessed to have an unclear RoB, owing primarily to the methodological characteristics. Imprecision because of the limited number of animals employed and inconsistent findings further increased the uncertainty of evidence.

In addition, it must be noted that the data included for analysis in this systematic review were drawn from animal experiments and cannot be directly extrapolated to humans. Also, the studies involved administration of pharmacological substances for periods of time different from those used commonly in clinical settings¹⁴  and dosages without rationale calculation for equivalence to humans.⁴⁴  The generalizability of results presented in this synthesis may be limited by the methodology for assessment of relapse; most of the studies investigated relapse following tipping tooth movement for varying periods of time.

OR continues to be difficult to manage for most clinicians, hence further research is warranted. In keeping with the existing knowledge on the biological basis of OR,^6–8  pharmacological agents with antiosteoclast effect show promise in regulating OR. Future research must explore drugs belonging to this class. It is also imperative that future animal studies follow standardized protocols,⁴⁵  with consideration to simulate human clinical scenarios with regard to the duration of drug administration, dose equivalence, and route as well as the attributes of mechanisms inducing tooth movement and methodology of relapse assessment. Emphasis must also be given to sample size calculation to enhance the power of the study and the significance of the results.⁴⁶ 

• Additional evidence of higher quality is required to draw definitive conclusions on the effects of the investigated pharmacological agents on tooth movement relapse following cessation of orthodontic force. Notwithstanding the fact that the overall certainty in evidence inhibits potential recommendations for human trials, the findings of this review provide a direction for future research.