Knowledge of orthodontic tooth movement through the maxillary sinus: a systematic review

Background To investigate the feasibility, safety and stability of current interventions for moving teeth through the maxillary sinus (MTTMS) by performing a systematic review of the literature. Methods The electronic databases PubMed, Embase, CENTRAL, Web of Science, CBM, CNKI and SIGLE were searched without a language restriction. The primary outcomes were parameters related to orthodontic treatment, including orthodontic protocols, magnitude of forces, type of tooth movement, duration and rate of tooth movement, and remolding of alveolar bone and the maxillary sinus floor. The secondary outcomes were safety and stability, including root resorption, perforation of the sinus floor, loss of pulp vitality and periodontal health and relapse. Results Nine case reports with 25 teeth were included and systematically analyzed. Fifty to two hundred g of force was applied to move teeth through the maxillary sinus. Bodily movement was accomplished, but initial tipping was observed in 7 cases. The rate was 0.6–0.7 mm/month for molar intrusion and 0.16–1.17 and 0.05–0.16 mm/month for mesial-distal movement of premolars and molars, respectively. Bone formation and remolding of the sinus floor occurred in 7 cases. Root resorption within 6 to 30 months was observed in 3 cases, while no cases of perforation of the sinus floor, loss of pulp vitality, periodontal health impairment or relapse were reported. Conclusions At the present stage, no evidence-based protocol could be recommended to guide MTTMS. The empirical application of constant and light to moderate forces (by TAD, segment and multibrackets) to slowly move teeth through or into the maxillary sinus in adults appears to be practical and secure. Bodily movement was accomplished, but teeth appear to be easily tipped initially, potentially resulting in root resorption. However, this conclusion should be interpreted with caution as the currently available evidence is based on only a few case reports or case series and longitudinal or controlled studies are lacking in this area.


Background
With the development of digitalization and material science in the past few decades, substantial progress has been achieved in orthodontic techniques for more efficient, precise, invisible, comfortable and rapid treatment [1][2][3]. However, orthodontists commonly encounter predicaments related to dental status, periodontal status, general health, orthodontic technique, anchorage, and other factors that may limit orthodontic treatment [4].
Among these challenges, movement of teeth against anatomic structures, such as the maxillary sinus (MS), appears to be non-evidence-based.
The MS is the largest paranasal sinus, located in the posterior maxilla, and has a close relationship with adjacent structures. It sprouts late in fetal life, existing at birth with a dimension of approximately 3*6*8 mm, and ends its growth in a pyramid shape in adults [5][6][7]. The MS floor (MSF) consists of a thin bony plate covered with a layer of mucosa. With pneumatization and aging, the floor extends into the posterior alveolar process and forms the alveolar recess, creating protrusions of root apices into the sinus [5]. Generally, the MSF is at the level of nasal floor at puberty and reaches its lowest point with the eruption of the third molars [5,6]. Morphologically, the sinus-root relationship (SRR) can fall into 5 categories: 0, the root is not in contact with the sinus floor; 1, the MSF curves inferiorly with the root in contact with the MSF; 2, the MSF curves inferiorly with the root projecting laterally on the sinus cavity but its apex is outside the sinus; 3, the MSF curves inferiorly with the root apex projecting on the sinus cavity; and 4, the MSF curves superiorly with part or all of the tooth root enveloped [8]. For individuals with excessive pneumatization, a type-1, 2, 3 or 4 relationship can be common.
The classic theory of orthodontic tooth movement stresses the dynamic balance of bone resorption on the pressure side and deposition on the tension side of the periodontal ligament (PDL) [9]. This theory has been successfully applied by orthodontists to move teeth in the alveolar bone. However, applying this concept to the MS, with potential tooth movement against cortex or soft tissue [10], is more difficult. Consequently, clinicians often fear the uncertainty of moving teeth through the maxillary sinus (MTTMS). However, recent experiments have revealed a particular biomechanical pattern regarding MTTMS. Mechanical stress could induce osteogenesis on the sinus side before bone resorption occurred on the PDL side, and the bone thickness of the sinus wall could be maintained [11][12][13], potentially indicating the feasibility of MTTMS. Furthermore, concurrent root resorption and higher efficiency of light forces were also observed in these experiments [11][12][13].
In orthodontic clinics, clinicians may encounter MTTMS. In some cases, planning the distalization of molars or the maxillary dentition to correct type-II occlusion or to achieve a better anterior profile is preferred because this technique has the benefit of avoiding extraction and is reported to be one of the advantages of clear aligners [1,9,14]. In other cases, when closure of posterior spaces [15][16][17], tooth intrusion to create spaces for opposing prosthetics [18,19] or an alternative non-surgical sinus lift for implant sites [20][21][22] is required, orthodontists must implement MTTMS. MTTMS determines the feasibility, duration and quality of comprehensive treatment. No systematic reviews on MTTMS are currently available. The purpose of this research was to systematically review the literature and investigate the feasibility, safety and stability of current interventions for MTTMS.

Methods
This systematic review was conducted generally following the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) checklist [23]. The literature search, data extraction and quality assessment were all performed independently by two reviewers. Any disagreement was resolved by discussion or by consultation with a third party.

Inclusion criteria
We set the following inclusion criteria to identify eligible studies: (1) Patients with at least one target tooth, defined as a tooth with at least one root protruding into the MS, were investigated. The morphological SRR (type-2, 3 or 4 for distal-mesial movement and type-1, 2, 3 or 4 for intrusion) must be confirmed by radiological diagnosis: periapical films, panoramics or cone-beam computed tomography (CBCT); (2) An orthodontic treatment to move target teeth through the MS was executed; (3) The primary outcomes were parameters related to orthodontic treatment, including orthodontic protocols, magnitude of orthodontic forces, tooth movement type, duration and rate of tooth movement, and remolding of the alveolar bone and MSF. The secondary outcomes were safety and stability, including orthodontically induced root resorption (OIRR), perforation of the sinus floor (Perforation of the sinus floor should be verified by occurrence of sinusitis or by radiological findings in sinus. The integrity of lamina dura should be assessed by radiography: periapical films, panoramics or CBCT, while the sinus membrane should be assessed with CT/CBCT, MRI or endoscopy [5,18,19,[24][25][26][27][28].), pulp vitality loss, and periodontal health impairment and relapse; and (4) The study was a clinical study, including randomized clinical trial, controlled clinical trial, cohort study, case-control study, cross-sectional study and case report.

Search strategy
Online searches were conducted in electronic databases, including PubMed, Embase, CENTRAL, Web of Science, Chinese Biomedical Literature Database (CBM), China National Knowledge Infrastructure (CNKI), without restriction of language. Grey literature was searched in the System for Information on Grey Literature in Europe (SIGLE). We used MeSH terms as well as free text words, and the key words were "maxillary sinus," "orthodontics," "orthodontic*," "tooth moving," and "tooth movement" for all databases. The reference lists of relevant articles were manually searched for additional studies. The searches were initially conducted in January 2017 and were updated on May 16, 2017.The search strategies are presented in Table 1.

Data extraction and analysis
Information regarding the characteristics and outcomes of the included studies was extracted. Specifically, the following characteristics were identified: country, age, sex, sample size, target teeth, SRR, source of active force and radiological method. The outcomes were those items defined in the inclusion criteria above.

Characteristics of the included studies
The online search yielded 677 articles. After excluding duplicate and irrelevant articles by reading titles and abstracts, 11 full texts remained. Then, the reference lists of these articles were read and one additional study was identified. No prospective or retrospective controlled clinical studies were found. Nine case reports meeting the inclusion criteria were included and systematically analyzed (Fig. 1). The characteristics of the 9 case reports are presented in Table 2.

Results of quality assessment
All included studies were case reports and consequently had a relatively high risk of bias.

Study outcomes analysis
In total, nine adult subjects with 25 target teeth were included. All target teeth had a type-2, 3 or 4 SRR (Table 3).

Protocol and magnitude of forces
Cacciafesta et al. [17] used segments to protract tooth numbers 27 and 28 mesially, and the force was 50 g by coil spring. Re et al. [22] used an endosseous implant in the retromolar area and a T-loop to move tooth number 25 distally, and the active load was 50 g/mm. In Kravitz et al.'s article [18], tooth number 16 was intruded using temporary anchorage devices (TADs), and the forces were 100 g by elastic power chain in the initial 2 months and 150 g by coil spring in the next 4 months. Yao et al. [19] also used TADs to intrude two adjacent molars (26 and 27), and the forces were 150-200 g by elastic power chain. Kuroda et al. [14] performed group distalization of the maxillary dentition using multibrackets and TADs, and 9 teeth were moved distally through the MS bilaterally with a load of 200 g by coil spring. Oh et al. [16], Park et al. [15], Saglam et al. [21] and Carvalho et al. [20] used multibrackets to move maxillary teeth mesially or distally through the MS (TADs were utilized in Park et al.'s article). In these reports, "light forces" or "mild to moderate forces" were used. Generally, light to moderate forces (50-200 g) were applied to accomplish MTTMS.

Alveolar bone formation and remolding of the sinus floor
Re et al. [22], Saglam et al. [21], and Carvalho et al. [20] moved the second premolars distally through the MS with pneumatization into the alveolar bone. Alveolar bone formation occurred in the moving direction, along with direct remolding of the sinus lamina dura and sinus lift, and implants were subsequently inserted in the previous positions of the second premolars. Likewise, alveolar bone formation was observed in the studies of Cacciafesta et al. [17] and Oh et al. [16], and signs of sinus wall modeling were also observed in Oh et al.'s    [18,19].

Stability and relapse
Oh et al. [16], Park et al. [15] and Kuroda et al. [14] reported stable occlusion after follow-ups of 1.5, 1 and 5 years, respectively. The periodontal health and vitality of the teeth were maintained in Yao et al.'s case [19]. Saglam et al. reported acceptable intraoral tissue health after 2 years [21].

Discussion
This systematic review intended to determine the feasibility, safety and stability of current interventions for MTTMS. Nine case reports representing the available human clinical studies were included. In general, the present study indicated the feasibility of MTTMS. However, the difficulty of the moving process varied substantially, possibly indicating the heterogeneity among clinical measures and internal anatomic structures and the inherent bias of case reports. In MTTMS, bodily movement is desired. The key biomechanical objective is uniform distribution of orthodontic forces along the PDL and the line of the active force passing through the center of resistance [29,30]. Carefully designed segments or TADs can produce approximate determinate force systems and may facilitate bodily movement [9,17,31,32]. Considering the anatomic variability of the MSF and the complexity of the SRR [8], techniques for better control in three dimensions should be developed, especially for patients with primarily regional complaints. To achieve tooth movement by frontal resorption, mild and constant forces (35-60 g, 70-120 g and 10-20 g for tipping, bodily and intrusion movement, respectively) are recommended. However, considering the amount of resistance in sliding mechanics [9,30], the decay rate of forces and the root numbers of the posterior teeth, the magnitude of 50-200 g of force applied in the included cases seems safe.
In the present study, most cases showed initial stages of tipping through the MS, which is consistent with a previous study [33]. Deviation from ideal bodily movement may reflect expression of a well-designed pure Newtonian mathematical force system applied on the in vivo PDL. Orthodontic forces are derived from deformation of some parts of existing appliances; however, each appliance has a particular load deflection rate, and the decay rates of the counterparts (i.e., the moment of force and the moment of couple) in the equilibrium system are not equal, and consequently, the moment to force ratio constantly changes, resulting in constant changes in the center of rotation and difficulty in maintaining translation [34]. Moreover, in the MS, the distribution of bone density along the axis of a tooth must be considered. The coronal part of the root is more likely to move against cancellous bone while the apical part is more likely to move against cortical bone [9]. Therefore, the tooth is easily tipped toward the moving direction. Furthermore, for molar intrusion, the accompanying tipping may reflect different resistances among roots [19].
Moving at low speeds is a prerequisite for compensatory bone regeneration [16,35]. Therefore, applying light and continuous forces is the best strategy to achieve the ideal speed for moving teeth. Particularly, the cask effect regarding the moving rate in MTTMS is probably due to cortical anchorage [9]. In Oh et al.'s case, the location of the roots against the cortical bone of the sinus wall provided nearly absolute anchorage in the first 3.5 years [16]. In such cases, only light forces are appropriate as heavy forces against the cortical bone will reinforce the anchorage and cause additional OIRR. However, individual heterogeneity may exist in this regard because no other authors reported such an extreme situation, and a wide range of rate was reported across cases.
According to the theory of tooth movement, new physiological bone along the course of the moved tooth can be harvested [9,36]. In the past few decades, the development of implant sites with the help of orthodontic tooth movement has been shown to be practical [36], and in several cases, this technique was successfully applied in the MS area [20][21][22]. One major concern, however, is maintaining the intact membrane of the sinus floor. First, in surgical sinus augmentation procedures, the floor is mechanically and instantaneously lifted by applying graft materials or alveolar bone blocks [37,38]. This shows the endurance and reparability of the sinus floor as it may adapt to slow and mild tooth movement. Second, recent studies have revealed that under stressful stimulation, bone deposition on the sinus side preceded resorption on the PDL side, and the amount of bone in the sinus wall was maintained or increased. This mechanism may partially account for bone remolding of the sinus floor [11,12].
Currently, the etiology of OIRR is unclear. However, comprehensive orthodontic treatment, particularly the application of heavy forces, undoubtedly cause increases in the incidence and severity of root resorption [39]. Moving a tooth against the cortical bone is more likely to cause heavy root resorption, but serious OIRR was not reported in any of the included cases, perhaps because of the light forces applied. However, lack of detection of serious OIRR in some cases may reflect a limitation of radiographic approaches [18,19,22]. On the one hand, there is hysteresis between histological and radiographical changes, and early root resorption could only be detected on radiographs after 6-12 months [39]. On the other hand, panoramics or periapical films were used in most cases [14,[17][18][19][20][21][22], but their diagnostic accuracy may be insufficient [39,40]. For retention, wrap-around or bonded retainers were mainly used for the maxillary teeth, and implants and subsequent prosthetic crowns were installed adjacent to or opposing the moved teeth. All these factors contributed to good retention and stability after MTTMS.

Limitations
Currently, the comprehension of MTTMS is limited. First, prospective controlled clinical trials with large samples are not available, so an optimal orthodontic protocol has not been established. Second, the techniques used to evaluate the SRR, OIRR and perforation of the sinus floor have low accuracy [41] and applying panoramics and periapical films can introduce errors in patient selection or outcome measurements. To improve accuracy, CBCT is an alternative strategy [8,24,25,40,42]. Third, the results of basic research simulating MTTMS are not necessarily authentic. Although the discovery of recent research was novel and was partly consistent with some clinical observations [11][12][13], the studies involved only a 2-week observation period on mouse models. And the results were not entirely consistent with the long-term findings in a previous biopsy report in human in which osteoclasts and obvious lamina dura resorption were observed on the sinus side [41]. Therefore, more basic studies with consistent models should be conducted to confirm these results. Lastly, during orthodontic tooth movement, some side effects such as severe root resorption, osseous perforation, and sinus perforation may be beyond orthodontists' control. These cannot be macroscopically or radiologically detected but can be verified histologically. Clinicians should determine accurate diagnoses with consideration of anatomical structures before treatment, execute careful protocols, and conduct progress evaluations throughout treatment [39][40][41].

Conclusion
At the present stage, no evidence-based protocol could be recommended to guide MTTMS. The empirical application of constant and light to moderate forces (by TAD, segment and multibrackets) to slowly move teeth through or into the maxillary sinus in adults appears to be practical and secure. Bodily movement could be accomplished, but teeth seem to be easily tipped initially, potentially resulting in root resorption. However, this conclusion should be interpreted with caution as the currently available evidence is based on only a few case reports or case series, and longitudinal or controlled studies are lacking in this area.

Funding
The National Natural Science Foundation of China (Nos. 81470722 and 81201379) supported the design of the study and collection, analysis, and interpretation of data and writing and publishing the manuscript.

Availability of data and materials
The summary of data extraction in this study is available upon request to the corresponding author.
Authors' contributions WS, KX, JL conducted the data search, extraction, assessment and the statistical analysis and draft the manuscript. XH made the figures and tables. XC, QL organized the structure of the manuscript and edited the language. WS and JL designed the study and revised the manuscript. All authors read and approved the final version of submission.
Ethics approval and consent to participate Not applicable.