In this study we investigated the relationships among the risk factors for inferior alveolar nerve injury (IANI), and the difference between preoperative imaging findings on panoramic radiographs and computed tomography (CT), by univariate and multivariate analyses. We determined the following to be significant variables by multivariate analysis: panoramic radiographic signs, such as the loss of the white line of the inferior alveolar canal or the diversion of the canal; excessive haemorrhage during extraction; and a close relationship of the roots to the IAN (type 1 cases) on CT examination. CT findings of type 1 were associated with a significantly higher risk (odds ratio 43.77) of IANI. In addition, many panoramic findings were not consistent with CT findings (275 of 440 teeth; 62.5%). These results suggest that CT findings may be able to predict the development of IANI more accurately than panoramic findings. Panoramic radiography alone did not provide sufficiently reliable images required for predicting IANI. Therefore, when the panoramic image is suggestive of a close relationship between the impacted tooth and the IAN, CT should be recommended as a means of conducting further investigations.
The removal of the mandibular third molars to prevent future problems is widely acknowledged to be useful and is one of the most commonly performed surgical procedures in oral surgery. The surgical removal of these molars may damage the nerve and cause hypoesthesia of the lower lip as an unpleasant postoperative complication. Many investigators have reported that inferior alveolar nerve injury (IANI) is associated with mandibular third molar removal. However, there have been few quantitative analyses of the factors contributing to IANI. In addition, other than our recent study, only a few studies to date have evaluated the multivariate relationships among the various risk factors and IANI. In our recent study, we reported some radiographic signs (the loss of the white line of the inferior alveolar canal, the diversion of the canal, and a close relationship of the roots to the IAN on panoramic radiographic findings and not on CT findings) to be significantly associated with IANI following third molar extraction.
Panoramic radiographs are the imaging modality most commonly used by oral and maxillofacial surgeons to view impacted third molars and to estimate the risk of IANI. In particular, Rood and Shehab reported that three of seven classic radiographic signs indicate a higher risk of IANI, implicating close proximity of the inferior alveolar nerve (IAN) to the lower third molar. However, some investigators have reported that panoramic radiography does not provide sufficiently reliable images to predict nerve lesions. In addition, some reports have indicated that it is only possible to determine the true relationship of the tooth root with the IAN with computed tomography (CT). In fact, axial, coronal, and sagittal CT images all provide surgeons with useful information, and such images are also beneficial for the preoperative planning of the surgical procedure because of the high-resolution quality of medical CT. In this study, we investigated the multivariate relationships among the various risk factors of IANI, and the differences between preoperative imaging findings on panoramic radiographs and medical CT images.
Patients and methods
This was a non-randomized, retrospective (historic) cohort study of patients. Thus, this study was granted exemption from institutional review board approval. Between April 2006 and March 2010, 2528 surgical removals of mandibular third molars were performed by dentists at the Department of Oral and Maxillofacial Surgery.
A total of 376 (high risk of IANI; types 1, 2, and 3 as described in ‘Study design’ section) of the 2528 teeth were found to be in close association with the mandibular canal on panoramic radiograph, or were expected to be very difficult to extract (for example, in cases with hypertrophic roots and/or dilacerated teeth). Sixty-four (low risk of IANI; types 4 and 5) of the 2528 teeth were opposite to the other teeth that were in close proximity to the IAN. A total of 440 teeth from 295 patients (122 men and 173 women) were included in this study. The patients had a mean age of 36.2 ± 12.2 years (range 16–71 years). Thus the data (total 440 teeth) for this study pertain to the impacted lower third molars considered to have a high level of risk of IANI during surgery. Before surgery, each patient was informed of the necessity to perform a CT examination and the possible complications, including the potential risk of IANI during the procedure.
Every intervention was carried out under local or general anaesthesia. The extraction of 147 teeth (33.4%) in 86 patients was carried out under general anaesthesia; for 293 teeth (66.6%) in 209 patients, the procedure was carried out under local anaesthesia. Sedation was not carried out. Envelope (sulcular) mucoperiosteal flaps were raised for superficial impactions, and triangular flaps were raised for deep impactions, followed by vestibular bone removal. If necessary, bone removal and crown and root sectioning were performed with tungsten fissure burs (Dentsply-Maillefer, Ballaigues, Switzerland). Sockets were irrigated with 20 ml sterile saline solution at room temperature, and exposure of the IAN was checked during and after precisely focused careful suction. A 3–0 silk suture was used to close the wound. An antibiotic and an anti-inflammatory drug were prescribed (usually oral cefcapene pivoxil hydrochloride hydrate 900 mg 3 times daily for 3 days and oral acetaminophen 1200 mg 3 times daily for 3 days). After 7 days, a surgeon removed the suture and explored the wound for closure and recorded any complications such as ecchymosis, trismus, swelling, postoperative bleeding, dry socket, infection, and IANI.
Patients who underwent conventional panoramic radiography (Orthoceph OC100CT; Yoshida, Tokyo, Japan) were included in the study. The radiation dose was 20–30 mSv. The medical CT images were taken using a High-Speed Advantage CT Scanner (GE Medical Systems-Japan, Tokyo, Japan). The exposure factors were set at 120 kV and 200 mA, and the duration of scanning was 6 s. The slice thickness of contiguous sections was 0.625 mm. Axial planes were set parallel to the occlusal plane, and continuous 0.625-mm slices were taken. Coronal and sagittal images were reconstructed from the raw data. The software program used was Advantage Workstation ® (GE Medical Systems-Japan). The radiation dose was 2.0–3.0 mSv.
The primary predictive variable was the presence or absence of one or more panoramic radiographic or CT findings. The panoramic radiographic and CT findings were analyzed independently by the first and second authors. These two authors were blinded independently to the postoperative IANI status. The CT images were presented to the two observers in a randomized order to prevent referral to the panoramic findings. Whenever a disagreement occurred between the two observers’ assessments, a consensus was reached by discussion. At first, the panoramic findings were evaluated for the presence or absence of the following three classic radiographic signs, all of which have been reported as suggestive of a close proximity of the IAN to the lower third molar : (1) darkening of the root where it crosses the inferior alveolar canal (as described below, this group was classified into types 1 and 2 on panoramic radiographic findings and not on CT findings); (2) interruption or obliteration of either of the radiopaque white (cortical) lines of the inferior alveolar canal; and (3) diversion or bending of the inferior alveolar canal in the region of the root apices. In addition, the relationship of the roots to the IAN in the panoramic radiographic findings was diagnosed radiographically in accordance with the method reported by Tanaka et al. ( Fig. 1 ). In type 1 cases, the canal is superimposed over more than half of the root structure. In type 2 cases, the canal is superimposed over less than half of the root structure. In type 3 cases, the root structure impinges on the superior border of the canal. In type 4 cases, the distance between the root tip and the superior border of the canal is less than 2 mm, and in type 5 cases, the distance between the root tip and the superior border of the canal is more than 2 mm.
The relationship of the roots to the IAN was diagnosed by CT with the same method that was used for the panoramic radiographic findings ( Fig. 2 ). In types 1 and 2, the disappearance of cortication of the canal is observed. In type 1, displacement of the nerve contents to curve around the root is observed. In the CT images, the relationship was also classified in terms of the buccolingual position into one of the four following categories: buccal, inferior, lingual, or inter-radicular ( Fig. 3 ).
The following were analyzed and compared by the method reported in our previous study : preoperative radiographic findings, such as the type of impaction (according to Pell and Gregory ), angulation (Winter classification ), three radiographic signs associated with IANI (listed above), and the classification (types 1–5) according to the method reported by Tanaka et al. ( Fig. 1 ) ; procedure-related complications during surgery, such as the observation of the canal, excessive haemorrhage, and bone removal; and demographic factors, including patient age and gender. All factors are listed in Table 1 .
|Variables||IANI present, n (%)||IANI absent, n (%)||P -value|
|Sample size||28 (6.4)||412 (93.6)|
|Male||14 (53.8)||108 (40.1)|
|Female||12 (46.2)||161 (59.9)||NS|
|Mean ± SD||40.5 ± 12.4||35.4 ± 11.9||<0.05 a|
|Right vs. left|
|Right||14 (50.0)||199 (48.3)|
|Left||14 (50.0)||213 (51.7)||NS|
|Vertical||4 (14.3)||73 (17.7)|
|Horizontal||9 (32.1)||132 (32.0)|
|Mesioangular||13 (46.4)||145 (35.2)|
|Distoangular||1 (3.6)||21 (5.1)|
|Transverse||1 (3.6)||41 (10.0)||NS|
|Pell and Gregory|
|I||1 (3.6)||104 (25.2)|
|II||16 (57.1)||226 (54.9)|
|III||11 (39.3)||82 (19.9)||<0.05 b|
|A||3 (10.7)||120 (29.1)|
|B||19 (67.9)||220 (53.4)|
|C||6 (21.4)||72 (17.5)||NS|
|Extent of root tip–inferior alveolar canal overlap (the method of Tanaka et al.) ( Fig. 1 )|
|Type 1||24 (85.7)||149 (36.2)|
|Type 2||3 (10.7)||90 (21.8)|
|Type 3||1 (3.6)||109 (26.5)|
|Type 4||0 (0)||38 (9.2)|
|Type 5||0 (0)||26 (6.3)||<0.05 b|
|Signs indicating close spatial relationship|
|Loss of the white line|
|Yes||9 (32.1)||13 (3.2)|
|No||19 (67.9)||399 (96.8)||<0.05 b|
|Diversion of the canal|
|Yes||11 (39.3)||17 (4.1)|
|No||17 (60.7)||395 (95.9)||<0.05 b|
|Extent of root tip–inferior alveolar canal overlap (the method of Tanaka et al.) ( Fig. 3 )|
|Type 1||26 (92.9)||60 (14.6)|
|Type 2||2 (7.1)||65 (15.8)|
|Type 3||0 (0)||127 (30.8)|
|Type 4||0 (0)||81 (19.7)|
|Type 5||0 (0)||79 (19.2)||<0.05 b|
|Signs indicating close spatial relationship|
|Buccal position||3 (10.7)||143 (34.7)|
|Inferior position||0 (0)||195 (47.3)|
|Lingual position||21 (75.0)||74 (18.0)|
|Inter-radicular position||4 (14.3)||0 (0)||<0.05 b|
|Yes||5 (17.9)||12 (2.9)|
|No||23 (82.1)||400 (97.1)||<0.05 b|
|Yes||5 (17.9)||1 (0.2)|
|No||23 (82.1)||411 (99.8)||<0.05 b|
|Yes||28 (100.0)||325 (78.9)|
|No||0 (0)||87 (21.1)||<0.05 b|
|1–4 years||3 (10.7)||68 (16.5)|
|5–9 years||14 (50.0)||161 (39.1)|
|>10 years||11 (39.3)||183 (44.4)||NS|