Distance and Direction of Retained Root Migration After Lower Third Molar Coronectomy

Background: The present study evaluated the pattern of root migration following coronectomy of the mandibular third molar in terms of distance, degree of direction, and relevant factors related to root migration. Methods: This retrospective study included 50 coronectomies in 44 patients with at least 1-year follow-up. Panoramic radiographs were taken pre-operatively, within 2 weeks after surgery, and at 3, 6, and 12 months post-operatively. Multiple factors are possibly related to root migration, so we analyzed sex, age, tooth, gure of root, residual bone height, and also as Winter’s and Pell& Gregory classication with respect to angulation, class, and position of tooth. Results: Over the rst three months after the coronectomies, all retained roots moved and/or changed their root axis. The respective mean distance of retained root migration in the horizontal (C1), coronal (C2), and oblique (C3) direction during the rst year post-operatively was 3.14 ± 1.86 mm, 2.42 ± 1.61 mm, and 3.45 ±1.76 mm. The mean (±SD) root axis change was 11.26 ± 5.55 degrees. The signicant inuencing factors related to root migration were sex, age, and in particular tooth angulation (GEE: P<0.05). The mesio- and horizontal angulation (M, H) group migrated signicantly further horizontally forward (C1) over time than the vertical (V) and distoangulation (D) group. The V, D group showed signicantly greater coronal migration (C2) (P=0.05) than the M, H group. The V, D group had only mesial rotation. Conclusion: These ndings could contribute to evaluation and/or planning for root removal. current investigated the degree and pattern of retained root movement following coronectomy and movement in direction using radiograph, and other Knowing the distance and direction of root and their for planning coronectomy and determining its prognosis.

A number of studies have addressed the magnitude of root migration with various measuring methods and reference points (1,10,12,24) . One study reported the pattern of root translation and rotation by using cone beam computed tomography (CBCT) 25 . Some researchers (1,10,12,25) reported on root migration distance and migration direction, while only one investigated the degree of root axis change to evaluate the directional change of the retained root (25) .
The current study investigated the degree and pattern of retained root movement following coronectomy in terms of distance and movement in any direction using panoramic radiograph, and other related factors. Knowing the distance and direction of retained root migration and their related factors would be useful for planning coronectomy and determining its prognosis. Patients were recruited who underwent coronectomy of an impacted lower third molar at the Oral and Maxillofacial Clinic, Faculty of Dentistry, Mahidol University, between January 2011 and January 2016, and who were at high-risk of IAN injury signs, according to Roods and Shehab's criteria 26,27 . The inclusion criteria were healthy patients who were followed up at least 1 year after the coronectomy and who had signed informed consent, had complete charting together with 5 periods of panoramic radiographs. The radiographs were pre-operative, immediate (within 2 weeks), and 3, 6, and 12 months post-operatively. If the patients had pathologic lesions near the tooth, they were excluded 1,3 .

Methods
We recorded patient demographics (sex, age), tooth status (side of tooth, angulation, classi cation, and position according to Winter's and Pell & Gregory's classi cation), gure of root, and residual bone height of the alveolar crest after coronectomy. Panoramic radiographs were taken with a Kodak CS 9000 Carestream radiographic machine (Carestream Health Inc. New York, USA). The retained root migration distance was measured in millimeters and the directional changes of the root axis were measured in degrees of angular rotation using PACS software (J.F. Advance Med Co., Ltd., Thailand)

Surgical technique
All lower third molar coronectomies were performed under local anesthesia as previously described 1-6, 9,10, 12-19,24 The tooth crown was removed and the resection surface trimmed to 4 mm below the alveolar crest to ensure no remaining enamel. Wound closure was done with Silk (Mersilk 3 − 0).
The X-and Y-axis were set to be the constant reference lines in all radiographs (Fig. 1A). The X-axis was the horizontal straight line passing both mesially and distally to the cementoenamel junctions (CEJs) of the lower second molar while the Y-axis was a vertical line perpendicular to the X-axis passing the distal CEJ of the lower second molar (A-point) As for root migration distance, the C point was set at the most inferior point of the retained root. To determine the root migration distance in the horizontal, coronal, and oblique directions, we measured the distance from the C point parallel to the X-axis (C 1 ) and the C point parallel to the Y-axis (C 2 ), and an oblique line from the C-point to the intersecting point of the X and Y axes (C 3 ). Measurements were in millimeters (mm).
The difference between C 1 , C 2 , and C 3 at time point T 0 , T 1 , T 2 , and T 3 (ΔCnTn) were calculated to represent the distance of root migration in the horizontal, coronal, and oblique directions at each time point. For example, ΔC 1 T 1 = C 1 T 0 -C 1 T 1 As for measurement of the direction of root migration (Fig. 1B), the root axis (M3 axis) was established from the E-point: the midpoint between the mesial and distal coronectomy resection margin (H 1 -H 2 ). The Changing in reference line angle were used to interpret the degree of direction to determine the tendency to mesial or distal rotation. The Y-axis served as a reference line for the mesio and horizontal angulation (M, H) impaction group (Fig. 1C), while the X axis was used for the vertical and distoangulation (V, D) impaction group (Fig. 1D). Increased degrees implied more mesial rotation while decreased degrees indicated more distal rotation of the roots.

Statistical analysis
Statistical analysis was performed using SPSS version 22 and Stata version 15. Inter-and intra-examiner reliability was assessed using the intra-class correlation coe cient (ICC). The distance and direction of root migration were reported using descriptive data analysis. The measured parameters were divided into groups, according to each potentially related factor. The difference between these groups were then examined using an independent T-test and one-way ANOVA. Based on these results, the Generalized Estimating Equation (GEE) was applied to the factors showing statistically signi cant differences so as to reveal the factors predictive for changes in distance and direction of root migration over time. P values < 0.05 were considered statistically signi cant. The Mann-Whitney U test was used to compare the rotational direction between the mesial and distal rotation groups.

Results
Fifty teeth from 44 patients were included. Patient demographics are presented in Table 1. Since only a small number of samples were available in each type of tooth angulation, two groups were formed as per similar characteristics of each impacted tooth: the M, H group, and the V, D group. The level of intra-and inter-examiner reliability was excellent with an ICC of 0.996 and 0.988, respectively. Root migration presented in all cases during the rst year after surgery. Root migration distances in 3 directions and the angular rotation of the root axis at each time point are presented in Table 2. The average root migration at the 1-year follow-up in the horizontal direction (mean ΔC1T3 ± SD) was 3.14 ± 1.86 mm, while it was 2.42 ± 1.61 mm in the coronal direction (mean ΔC2T3 ± SD), and 3.45 ± 1.76 mm in the oblique direction (mean ΔC3T3 ± SD). The average angular rotation (mean ΔâT3 ± SD) was 11.26 ± 5.55 degrees.

Factors related to root migration
The Independent t-test and one-way ANOVA of factors relating to root migration at the 1-year follow-up are presented in Table 3. Tooth angulation was statistically related to both migration distances (ΔC1, ΔC2) and angular rotation (Δâ) while age and sex were related to coronal migration (ΔC2) and angular rotation (Δâ), respectively (P < 0.05).  Although the shape of the root proved not to be statistically signi cant, it was considered in the GEE analysis because previous studies suggested it might in uence root migration 10,13,20,25 .
An assessment of all related factors in terms of distance in three directions (C1, C2, and C3), and angular rotation of the root axis revealed the progression of retained root migration followed similar patterns (Fig. 2-5). Peak migration rates of the root were found at 3-6 months then gradually decreased between 6-12 months post-operatively. According to the GEE analysis, only acceleration of horizontal (C1) and coronal (C2) root migration were signi cantly affected by tooth angulation.
The M, H group showed more horizontal migration (C1) than the V, D group over time (P = 0.003) (Fig. 2C) while the V, D group roots migrated more coronally (C2) than those in the M, H group (P = 0.05) (Fig. 3C). There was no signi cant relationship with any factors for oblique migration (C3) (Fig. 4) and angle changing (Fig. 5) (â). With respect to mesial and distal rotation, the changes in angular rotation were not statistically signi cant different between groups (Table 4).

Discussion
The results of the current study agreed with previous studies albeit follow-up times were different 1,10,12,13,20,25 . Goto et al. 10  The current study showed an acceleration of root migration, coronally in the V, D group, and horizontally in the M, H group, up to 1 year post-operatively. These results agree with Leung and Cheung (20) and Yeung et al. (25) who reported predominantly mesial translation or towards the oral cavity over time which should be followed-up for more than 6 months. (7,20) More than half of the roots in the M, H group were mesially-rotated, which should be a warning that the moving root may toward reimpacting to the second molar.
The distally-rotated roots were solely in the M, H group, suggesting an up-righting behavior of these roots, making them easier to be removed later. Furthermore, no roots in the V, D group had distal rotation, which would increase distal impaction of the root into the ramus region.
At a rapid migration rate, coronal and oblique movement of retained roots could emerge in the alveolar bone or be intraorally exposed within 1 year post-operatively. The V, D group migrated coronally a greater distance than the M, H group, which may be result in a greater chance of the retained root erupting into the oral cavities. Monaco et al. (12) reported that within the rst year, 6% of those migrated root fragments need to be removed. If they were not removed, the consequence of horizontal and oblique migration was likely to re-impact the adjacent second molar. The M, H group had signi cant horizontal migration, perhaps because the M, H group is most likely to migrate in the original direction towards the lower second molar root.
Kohara et al. (13) reported a signi cant difference in root migration between patients of different ages and sex. The current study suggests the signi cant difference in coronal migration and angular rotation is primarily related to age and sex, respectively.
Taken together, our results indicate that coronectomy helps to avoid IANI and if a second operation is needed to remove retained root, it is a less complicated surgery than the original total removal. Postoperative follow-up is vital, especially during on-going root migration. Surgeons should be aware of the consequences of a retained root abutting the lower second molar root which can develop a deep periodontal pocket and dental caries at the distal surface of the lower second molar. By way of prevention, the retained roots should be removed when they migrate away from inferior alveolar canal. This procedure should be performed before the root migrates to touching the second molars.
In cases of a partially or fully erupted V, D impacted tooth (position A), where the crown resection level might be lower than usual, it is necessary to compensate for the migration distance so as to reduce the need for re-operation. Notwithstanding, to our knowledge no studies describe the effects on the lower second molar. Further studies are thus needed to improve surgical techniques.

Conclusion
All retained roots of coronectomies changed in location and direction up to 1 year post-operatively. The in uencing factors were sex, age, and especially angulation of the tooth. There was signi cant acceleration of horizontal root migration in the M, H group compared to the V, D group. The latter trended to move more coronally. The V, D group change root migration direction only mesially and to a much greater degree than the M, H group.
These results could assist in managing impacted lower third molars at high risk of inferior alveolar nerve injuries with respect to surgical plans and information communicated to patients. The patient consent was not required in this study.

Abbreviations
Availability of data and material: The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.