External apical root resorption and vectors of orthodontic tooth movement


External apical root resorption is nearly ubiquitous in people treated orthodontically. This study predicted the extent of external apical root resorption by the vector of the incisor movement.


Cone-beam computed tomography scans of 93 white American adolescents (45 boys, 48 girls) with a Class I malocclusion who received comprehensive orthodontics were analyzed. Half were treated with 4 first-premolar extractions, and the others were treated without extractions. An x, y, z coordinate system was registered on the maxillae, superimposing on foramina, to quantify vectors of maxillary incisor movements. Multiple linear regression identified significant predictors of resorption for each incisor.


Strongly predictive models (R 2 = 77%-86%) were obtained. All directions of incisor movement tested (anteroposterior, mediolateral, craniocaudal, torquing) increased the risk of resorption in a dose-response fashion. Intrusion was most damaging. The patient’s sex, age, and duration of treatment were not predictive.


Root resorption is a very frequent consequence of tooth movement, especially intrusion and torquing, though no direction is harmless, and most corrections occur in combination. Incisor apical resorption was significantly greater in the extraction sample (ca 0.5 mm).


  • Root movement causes resorption, and more movement causes more resorption.

  • Directions of tooth movements incur different risks of apical root resorption.

  • Intrusion is particularly deleterious, whereas extrusion is comparatively benign.

  • Resorption may be inevitable until therapeutic prevention becomes available.

Root resorption is a frequent iatrogenic consequence of orthodontic therapy. Most of the time, external apical root resorption (EARR) does not affect tooth longevity, and net esthetic and functional benefits outweigh side effects. However, orthodontic tooth movement (OTM) heightens the risk and severity of root resorption, which is irreversible, as well as increasing risks of gingival change or color alteration. These are not novel findings, , but it awaited routine dental radiography before the scope of the problem became apparent. , Along with white-spot lesions, root resorption is the major iatrogenic consequence of orthodontic treatment.

There is no specific test to screen for those particularly at risk for EARR. Suggested indicators are (1) results of prior family members, (2) periodic radiographic monitoring, and (3) screening for a history of trauma and root abnormalities. In addition to genetic predisposition, the amount and types of OTM have been shown to affect the extent of EARR considerably. , , ,

Factors restricting pulp tissue respiration and promoting EARR are those that overcompress the periodontal ligament, leading to hyalinization. Roots are normally protected from dentinoclasia by unmineralized organic cementoid since clastic cells cannot adhere to unmineralized surfaces. Large overjet, long treatment, torquing, intrusion, and extraction treatment have been implicated. , The distance teeth are moved may be associated with the occurrence and severity of EARR. ,

Not only can the amount of movement affect the risk of EARR but also the direction. , For example, when compared with root extrusion, intrusion causes about 4 times more root resorption than extrusion. Guo et al concluded that a logical next step is testing whether the amount of root movement is a risk factor.

This project tested for associations between vectors of maxillary incisor movement and millimeters of EARR. The analysis was partly descriptive—distinguishing among directions of OTM—but differences of extraction, sex, age, and duration of treatment on the amounts of EARR were also tested.

Material and methods

Subjects were white American adolescents who received comprehensive orthodontic treatment with fixed appliances. Institutional approval was granted before study initiation (IRB #17-05545-XP). Patients were free of disease or any discernible syndrome and were fully dentate (no hypodontia, ignoring third molars). No patient with open incisor apices or treated with palatal expanders was included. Cases in which canines resorbed adjacent teeth were excluded from the study. Incisors were free of resorption (no apical “blunting”) by visual assessment of each cone-beam computed tomography (CBCT) at the start of treatment. All exhibited a Class I sagittal molar relationship at pre- and posttreatment examinations.

Half of the 93 patients were boys (45/93, 48%), and nearly half (43/93, 46%) were treated with 4 first-premolar extractions. All others were treated nonextraction. By 2-way analyses of variance, there was no significant difference by sex or treatment in starting or ending ages or in the duration of treatment. The average age at the start of treatment for the whole sample was 13.6 years (standard deviation [SD], 1.93), and at end of treatment, it averaged 15.4 years (SD, 1.93), with treatment duration averaging 1.76 years (SD, 0.57). The sample was from 1 multioffice group orthodontic practice, where patients were treated by 1 of 5 orthodontists rather than being assigned to 1 specialist.

All CBCTs were taken on iCAT Next Generation machines (Imaging Sciences, Hatfield, Pa) with settings of (120 kVp; 10.11 mA; 0.4-mm voxel size; scan time, 4.8 seconds; and field view of no more than 16 cm in height and 13 cm in depth). Patients stood upright in maximum intercuspation. CBCTs were taken for treatment and diagnostic purposes ( Fig 1 ). Voxel sizes of 0.2 mm and 0.4 mm have been shown to be equally useful for detecting EARR.

Fig 1
Examples of CBCTs, showing orientations using Anatomage. Lateral view (A) and frontal view (B) images were oriented in Frankfort horizontal with correction of any left-right cant.

All patients were treated using an MBT preadjusted appliance with 0.022-in slots. The general wire sequence was 0.014-in nickel-titanium followed by 0.018 × 0.018-in nickel-titanium or 0.018-in round nickel-titanium with coil springs as needed to create space. This treatment was followed by 0.017 × 0.025-in reverse-curve nickel-titanium M arch then 0.018 × 0.025-in nickel-titanium wires to finish. In extraction cases, space closure was typically performed on the 0.017 × 0.025-in reverse-curve nickel-titanium M arch with power chain and elastics (Class I, II, and/or III) as needed for anchorage control.

Multiple linear regression was used in a stepwise fashion to quantify the extent of EARR predicted from positional changes of each maxillary incisor during orthodontic treatment. We focused on the maxillary incisors because they are the teeth most prone to root resorption, , , perhaps owing to their single, cylindrical root, and because incisors commonly are moved more during treatment than posterior teeth. The outcome variable was root shortening with EARR being calculated on each maxillary incisor as a starting root length minus ending root length.

Two nominal predictors were tested, treatment (extraction or nonextraction) and patient’s sex (boy or girl), along with 8 ratio-scale variables: (1) mediolateral change in facial cementoenamel junction (CEJ) location, (2) anteroposterior change in CEJ, (3) craniocaudal change in CEJ, (4) the mediolateral change in incisor root apex position, (5) anteroposterior incisor apex movement, (6) craniocaudal apex movement, (7) amount of root torquing in the parasagittal plane, and (8) treatment duration. Factorial analysis of variance (ANOVA) was used to test group differences as indicated. Tests were 2-tailed and were evaluated at the conventional alpha level of 0.05.

Each CBCT was so situated that the left and right mandibular borders were superimposed in profile view. Coordinate axes ( Fig 2 ) were set such that the anteroposterior plane paralleled Frankfort horizontal, and the craniocaudal axis was at the facial midline in frontal view. Over the term of orthodontic treatment, we assumed that palatal growth was inconsequential. Because of software requirements, 4 palatal points were used for superimposition ( Fig 3 ), namely the left and right greater palatine foramina and the incisive fossa (posterior and anterior bony margins as the canal enters the oral cavity). The incisive fossa points were located in the axial view, whereas the palatine fossae were located in the coronal view. For each maxillary incisor, the facial CEJ at its most gingival extent and the root apex was located in the sagittal view. Landmarks were confirmed on the 3-dimensional rendering and in the other 2-dimensional views in InVivo (Anatomage, Inc, San Jose, Calif). Cartesian coordinates (x, y, and z) were recorded for each point as well as root length for each maxillary incisor, which was the straight-line distance from root apex to facial CEJ. One measure of method accuracy was that no root length change was positive from pre- to posttreatment.

Fig 2
A, Schematic diagram of a maxilla showing the 3 foramina ( stars ) used for superimposition between pre- and posttreatment CBCTs and the Cartesian arrangement: x-axis was mediolateral, y-axis was anteroposterior, and z-axis (not shown) was craniocaudal. Incisor root apex ( hexagon ) was identified, and fictitious locations are recorded before and after treatment. These coordinates were used to calculate tooth movement. B, Medial view of a maxillary central incisor showing the 3 landmarks, incisal edge, facial aspect of CEJ, and root apex. Incisal edge-to-CEJ distance often was reduced by cosmetic reshaping, but extents were unknown. The treatment change in CEJ-to-apex length measured millimeters of external root resorption (a negative number). C, Orientation of the 3-dimensional coordinate system, showing x- (mediolateral), y- (anteroposterior), and z- (craniocaudal) axes.

Fig 3
Close-up of the palate showing the foramina used for superimposition of the pre- and posttreatment CBCTs: A, midsagittal view of the incisive fossa. With the head oriented in Frankfort horizontal, anterior and posterior bony limits were located at the same height where the fossa emits into the oral tissue. B, In the frontal view, the left and right greater palatine foramen were located (lateral margin).

The straight-line distance of OTM for each incisor (using the CEJ) also was calculated:

<SPAN role=presentation tabIndex=0 id=MathJax-Element-1-Frame class=MathJax style="POSITION: relative" data-mathml='(x1−x2)2+(y2−y2)2+(z1−z2)2′>(?1?2)2+(?2?2)2+(?1?2)2(x1−x2)2+(y2−y2)2+(z1−z2)2

Straight-line distances are composites of x, y, and z changes and commonly explained more variance than single-direction vectors. Straight-line distances exceeded the variances accounted for by the individual vectors; therefore, they were excluded from the predictive models. That is, models using straight-line distances showed that OTM causes EARR, which is a trivial finding.

Torquing was included because an incisor root can be tipped appreciably through cancellous bone. An approximate formula was used:

<SPAN role=presentation tabIndex=0 id=MathJax-Element-2-Frame class=MathJax style="POSITION: relative" data-mathml='(y1CEJ−y1Apex)−(y2CEJ−y2Apex)’>(?1????1????)(?2????2????)(y1CEJ−y1Apex)−(y2CEJ−y2Apex)
Only gold members can continue reading. Log In or Register to continue

Jan 9, 2021 | Posted by in Orthodontics | Comments Off on External apical root resorption and vectors of orthodontic tooth movement
Premium Wordpress Themes by UFO Themes