We aimed to investigate the incidence and severity of orthodontically induced inflammatory root resorption (OIIRR) on maxillary incisors with clear aligner therapy using cone-beam computed tomography and to identify possible risk factors.
The root lengths of maxillary incisors were measured on orthogonal images from pretreatment and posttreatment cone-beam computed tomography examinations of 160 patients who received comprehensive orthodontic treatment with clear aligners.
Mean absolute reductions in root length varied between 0.47 ± 0.61 mm and 0.55 ± 0.70 mm and were not significantly different between maxillary central and lateral incisors. The prevalence of severe OIIRR, defined as both maxillary central incisors experiencing greater than a 25% reduction in root length, was found to be 1.25%. Potential risk factors included sex, malocclusion, crowding, and posttreatment approximation of apices to the cortical plates. Race, interproximal reduction, previous trauma to the teeth, elastics, age, treatment duration, and pretreatment approximation of apices to the cortical plates did not significantly affect the amount of OIIRR.
Comprehensive treatment with clear aligners resulted in minimal root resorption. Sex, malocclusion, crowding, and posttreatment approximation to the cortical plates significantly affected the percentage of change in root length. Posttreatment approximation of root apices to the palatal cortical plate showed the strongest association for increased OIIRR.
We investigated the incidence of root resorption in patients treated with clear aligners.
Comprehensive treatment with clear aligners resulted in minimal root resorption.
Mean resorption did not differ significantly between central and lateral incisors.
Risk factors for resorption were identified.
Proximity of apices to the palatal cortical plate showed the strongest association.
Orthodontically induced inflammatory root resorption (OIIRR) is a well-documented and common consequence of fixed appliance orthodontic therapy. The prevalence of severe root resorption, defined as loss of root length greater than 25% on both maxillary central incisors as shown on panoramic images, is reported to be 3%. Using cone-beam computed tomography (CBCT) images, Lund et al found that after fixed appliance therapy, 94% of patients had at least 1 root with greater than 1 mm of shortening, and 6.6% of patients had at least 1 root with resorption exceeding 4 mm. Other authors have found mean root resorption for the maxillary incisors to be 1.4 mm. Results from authors attempting to identify possible risk factors associated with OIIRR are equivocal due to variability in study design and sometimes contradictory conclusions. Many authors have agreed that maxillary central and lateral incisors are the most frequently resorbed teeth, but factors such as sex, duration of treatment, previous trauma, root morphology, certain features of malocclusion, and adjunctive treatments have little consensus as to their bearing on OIIRR. More recently, the possible influence of genetics on OIIRR has been investigated and found to contribute to the severity of root resorption, with interleukin-6 and interleukin-1B specifically linked to increased risks of root resorption.
Numerous authors have investigated the incidence and severity of OIIRR in patients who underwent fixed orthodontic therapy. Linge and Linge reported that fixed appliances resulted in more root resorption than did removable appliances. Barbagallo et al examined the effects of buccally directed force on premolars before their extraction, using microcomputed tomography to quantify the amount of root resorption. The results demonstrated that heavy-force (225 g) application resulted in a 9-fold increase in OIIRR compared with the controls (0 g), whereas light-force (25 g) application and thermoplastic aligners resulted in 5-fold and 6-fold increases in OIIRR, respectively.
Clear aligner therapy, a system of sequential, removable, extremely thin full-dentition plastic oral appliances, is designed to deliver small amounts of orthodontic movement to the teeth per aligner, so it is no surprise that there are anecdotal claims of minimal root resorption. Although clear aligners should theoretically result in less root resorption given their purported ability to minimize force and achieve more specific tooth movements, the limited published data on this orthodontic treatment modality suggests that patients undergoing treatment with clear aligners may still be susceptible to OIIRR. Brezniak and Wasserstein reported severe root resorption after treatment with aligners; measuring root-crown ratios on pretreatment and posttreatment panoramic radiographs, Kreiger et al found that 6.31% of teeth showed greater than a 20% reduction in root-crown ratio.
Most studies that have evaluated OIIRR for fixed appliance or clear aligner orthodontic therapy have relied on either panoramic or intraoral radiography to measure root resorption. The main disadvantage of these methods is that they underestimate root length. Sherrad et al reported that periapical radiographs underestimate root length on average by 2.6 mm compared with less than 0.3 mm for CBCT. Since many studies have relied on imaging techniques that tend to underestimate root resorption, it is plausible that the severity of root resorption for fixed appliances has been historically underreported. Additionally, panoramic radiographs can show distortion if teeth are positioned outside the focal trough as would be the case for anterior teeth in patients with Class II or Class III malocclusions. CBCT overcomes the limitations of panoramic and periapical radiographs, and studies show that with regard to assessing root length, CBCT results are highly reproducible and offer excellent sensitivity and specificity.
The purpose of this study was to use CBCT to investigate the incidence and severity of OIIRR in patients who had comprehensive treatment with clear aligners (Invisalign; Align Technologies, Santa Clara, Calif) and to identify possible risk factors associated with OIIRR.
Material and methods
This investigation was a retrospective experiment with pretreatment and posttreatment CBCT data from patients who had clear aligner therapy from 1 practitioner (D.G.). The study protocol was reviewed and approved by the Institutional Review Board Committee of the University of Louisville (number 16.0395).
The records of 458 patients who had consecutively finished clear aligner therapy by 1 orthodontic provider were reviewed for inclusion in the study. Inclusion criteria consisted of the records of patients who had diagnostically acceptable pretreatment and posttreatment records including photographs and initial and final CBCT data and who had comprehensive treatment with clear aligners. The sample consisted largely of nonextraction patients, with only 6 having extractions as part of their comprehensive treatment. Exclusion criteria included patients who had received previous orthodontic treatment, whether comprehensive or early interceptive treatment, and those treated with adjunctive appliances or a combination of clear aligners and fixed appliances. Of the 160 patients who met the criteria, all had pretreatment CBCT (T1) volumes taken within 6 months of beginning treatment. Most patients had posttreatment CBCT (T2) images made on the day that active treatment ceased; however, to limit patient exposure to unnecessary radiation, the orthodontic provider used the progress CBCT as the final image if that assessment was performed within 9 months of treatment completion.
CBCT images were acquired using an i-CAT device (Imaging Sciences International, Hatfield, Pa). It was operated at 3.8 mA (pulse mode) and 120 kV using a high frequency generator with fixed anode and a 0.5-mm nominal focal spot size. The anterior symphyseal region of the mandible of each patient was stabilized by a chin holder, and vertical and horizontal lasers were used to position the face. The patient’s head was oriented by adjustment of the chin support until the midsagittal plane was perpendicular to the floor, and the horizontal laser reference coincided with the base of the alar of the nose. Lateral scout images were made, and small adjustments to head position were carried out so that discrepancies between bilateral structures (eg, posterior and inferior borders of the mandibular rami and zygomatic arches) were less than 5 mm. A single 360° rotation, 20-second scan, comprising 306 basis projections was then made for each patient with a 17.0 cm (diameter) × 13.2 cm (height) field of view using the i-CAT acquisition software (version 1.7.7). Control of acquisition parameters, mA and kV(p), was automated. Primary reconstruction of the data was automatically performed immediately after acquisition and took approximately 60 seconds. The DICOM data from the CBCT scan was uploaded into 3-dimensional imaging software (Dolphin Imaging and Management Solutions, Chatsworth, Calif) for storage and interpretation.
DICOM data were exported anonymized from the Dolphin imaging software and codified according to the clinical and treatment parameters defined above. The DICOM data were imported into another dental software package (inVivo, version 5.3; Anatomage, San Jose, Calif) installed on an Alienware P18E laptop computer (Dell, Round Rock, Tex). All images were viewed on an 18.4-in WLED TrueLife display monitor screen (Dell) having a resolution of 1920 × 1080 pixels.
The CBCT volume was reoriented so that each of the orthogonal planes (axial, coronal, and sagittal) were perpendicular to the long axes of each maxillary central and lateral incisor under investigation ( Fig 1 ). A reference line was drawn using the proprietary software connecting the buccal and palatal cementoenamel junctions. The principal investigator (C.A.) measured the perpendicular distance between the most apical point of the tooth and the reference line at the cementoenamel junction ( Fig 2 ). The root lengths of the teeth in all pretreatment images were measured. After a minimum 3-week waiting period, posttreatment root length was assessed without access to the pretreatment data.
A minimum of 3 weeks later, the investigator repeated the reorientation protocol and root length measurements for T1 and T2 CBCT images for 10% of the sample by selecting every tenth patient in consecutive order. The rater was blinded to previous measurements.
Additionally, the pretreatment and posttreatment positions of the teeth relative to the orofacial cortical plates were qualitatively assessed according to the rating scale reported by Kan et al. The only modification was the addition of a fifth category that identified teeth outside the labial cortical plate. Each sagittal root projection in relationship to its orofacial cortical plates was classified as follows: Class I, the root is positioned against the labial cortical plate ( Fig 3 , A ); Class II, the root is centered in the middle of the alveolar housing without engaging either the labial or the palatal cortical plates at the apical third of the root ( Fig 3 , B ); Class III, the root is positioned against the palatal cortical plate ( Fig 3 , C ); Class IV, at least two thirds of the root is engaging both the labial and palatal cortical plates ( Fig 3 , D ); and Class V, the root is positioned outside the labial cortical plate ( Fig 3 , E ).
All data were entered into Excel (version 2017; Microsoft, Redmond, Wash). Statistical analysis was performed using SPSS software (version 23.0 for Windows; IBM, Armonk, NY). Intraclass correlation coefficients (ICC) were used to measure intrarater reliability for 2 repeated measurements. Demographics of the patients treated with clear aligner therapy, their clinical presentations, and treatment modifications were tallied. Mean tooth lengths at T1 and T2 were measured, and the percentage of root length change was calculated for all maxillary central and lateral incisors. The percentages of central and lateral incisors with greater than 25% root length reduction were identified. Repeated measures analysis of covariance was used to determine the effect of variables (discrete variables: tooth positions, sex, race, interproximal reduction, malocclusion, crowding, previous trauma to the teeth, use of elastics, and pretreatment and posttreatment approximations to cortical plates; continuous variables: age, treatment duration, and time elapsed between the T1 and T2 CBCT examinations) on the percentage of change in root length during orthodontic treatment. The Tukey LSD test was used for mean comparisons among the categories of each variable. The a priori level of significance was set at P = 0.05.
The records of 458 patients whose treatment with clear aligners was consecutively completed between July 2013 and August 2016 were reviewed; 160 records met the inclusion criteria. Table I shows the demographics of the subjects, and Table II shows their characteristics in regard to age at the start of treatment, treatment duration, and time interval between the T1 and T2 CBCT examinations. The mean age of the population studied was 34 ± 16 years. The mean total treatment time and mean time between the T1 and T2 CBCT evaluations were 2.19 ± 0.81 and 2.11 ± 0.85 years, respectively. Table III provides the frequency distribution for the dental characteristics of the sample and treatment details used to facilitate clear aligner therapy.
|Parameter||n||Mean ± SD|
|Age (y)||160||34 ± 16|
|Treatment duration (y)||160||2.19 ± 0.81|
|Interval between CBCT examinations (y)||160||2.11 ± 0.85|
|Interproximal reduction (yes)||101 (63)|
|Extraction (yes)||6 (4)|
|Elastics (yes)||82 (51)|
|Acceptable overbite (10%-40%)||69 (43)|
|Acceptable overjet (0-3 mm)||120 (75)|
|Open bite (yes)||16 (10)|
|Crossbite (yes)||27 (17)|
|Trauma (yes)||10 (6)|
|>Half-step Class II||30 (19)|
|<Half-step Class II||23 (14)|
|Class I||92 (58)|
|<Half-step Class III||7 (4)|
|>Half-step Class III||8 (5)|
|Mild (≤3 mm)||63 (39)|
|Moderate (4-7 mm)||49 (31)|
|Severe (>7 mm)||15 (9)|
Table IV shows the ICC used as an index of reliability of the rater for the 2 repeated measurements. The rater was highly reliable between the repeated measurements (ICC, 0.99; P < 0.01).
|Pretreatment||0.99 ( P <0.01)|
|Posttreatment||0.99 ( P <0.01)|
|Overall||0.99 ( P <0.01)|
Table V gives the mean reductions in root length between T1 and T2. Mean absolute reduction among maxillary incisors varied between 0.47 ± 0.61 mm (right lateral incisor) and 0.55 ± 0.70 mm (left central incisor). The range of change in root length was –5.24 to 0.23 mm. The mean percentage of reduction varied from 3.60% ± 4.50% to 4.30% ± 5.31%. Table VI shows a comparison between maxillary lateral and central incisor changes in root length. No differences were proven for root length reduction between central and lateral incisors when comparing either absolute differences in root length ( P = 0.55) or percentages of change ( P = 0.35). Table VII shows mean absolute and percentage changes and ranges in root length between T1 and T2 for each incisor. Applying a clinical threshold indicated that of the entire sample, only 2 or 3 subjects had greater than a 25% reduction in root length ( Fig 4 ).
|Maxillary incisor||Change in root length (mm)
(mean ± SD)
|Range (mm)||Change in root length (%)
(mean ± SD)
|Right lateral||−0.47 ± 0.61||−4.14-0.23||−3.60 ± 4.50||−27.64-1.80|
|Right central||−0.50 ± 0.61||−3.85-0.09||−4.03 ± 4.77||−27.56-0.85|
|Left central||−0.55 ± 0.70||−4.62-0.07||−4.30 ± 5.31||−32.67-0.70|
|Left lateral||−0.51 ± 0.70||-5.24-0.14||−3.87 ± 4.95||−32.87-1.05|
|Maxillary incisor||Change in root length (mm)
(mean ± SD)
|Range (mm)||Change in root length (%)
(mean ± SD)
|Laterals||−0.49 ± 0.57||−4.69-0.00||−3.72 ± 4.08||−30.25-0.00|
|Centrals||−0.53 ± 0.59||−4.24-0.04||−4.16 ± 4.50||−30.12-0.39|
|Right lateral||155 (99.36%)||1 (0.64%)|
|Right central||158 (98.75%)||2 (1.25%)|
|Left central||157 (98.13%)||3 (1.9%)|
|Left lateral||155 (99.36%)||1 (0.64%)|
|Both centrals||158 (98.75%)||2 (1.25%)|