This study investigated the exact intra-alveolar aetiology of a panoramic high-risk sign, darkening of the third molar roots. 83 mandibular third molar surgical removals demonstrating dark bands on the third molar roots in preoperative radiographs were included in this prospective study. Exposure of the inferior alveolar nerve (IAN), the root morphology of the third molar (e.g. groove or hook) and the integrity of the mandibular canal or lingual cortical wall were observed. Differences between single (increased radiolucency alone) and multiple darkening cases (increased radiolucency with accompanying ‘high risk’ signs) and between IAN exposure and groove formation were analysed. In 38 cases (45.8%), the IAN was visible during the operation. Groove was present in 37.4% of cases. 26.5% of the cases showed lingual cortical thinning, while specious root conformation explained the formation of darkening on the radiographic images of an additional 9.6% of the cases. IAN exposure ( P < 0.001) and groove formation ( P < 0.001) were significantly more frequent in multiple darkening cases than in single darkening cases. According to these findings, darkening of the third molar roots is more often the result of fenestration of the inferior alveolar canal wall or groove formation of the root than lingual cortical thinning.
Preoperative risk assessment of neurosensory disturbances before third molar surgery is essential. Panoramic radiography is widely accepted and used because of its advantages (cost effectiveness and low radiation levels). The values of different ‘high risk’ markers and specific signs have been evaluated and discussed to some extent. Darkening of third molar roots on panoramic radiographs was proved by several authors to be one of the strongest classic specific signs indicating a close anatomic relationship between the third molar roots and the inferior alveolar canal (IAC). In addition, Leung and Cheung stated that only darkening of the root was significantly related to postoperative inferior alveolar nerve (IAN) deficit.
Bundy et al. and Öhman et al. concluded that this radiolucent band is the result of root material loss (groove on the root) caused by the IAC, whereas Tantanapornkul et al. stated, according to their cone beam computed tomographic (CBCT) findings, that dark banding on panoramic radiographs is the sign of lingual cortical thinning.
The exact formation of this sign seems to be multi-causal. The results of the authors’ earlier investigations suggested a significant difference between so-called single and multiple appearances of dark bands on the third molar roots in relation to IAN exposure. Multiple darkening was defined as a dark band across the root with other simultaneous adjacent high risk signs (e.g. interruption of the superior cortical line of the canal and diversion of the canal or narrowing of the canal), and IAN exposure was seen significantly more frequently in these cases. Plenty of radiological investigations (including CT and CBCT) have dealt with image analysis of the different ‘high risk’ radiographic markers, including the darkening sign ; but direct visual information on intra-operative intra-alveolar findings and on root conformation connected to that sign is incomplete.
The authors’ aim was to examine the exact morphology of the third molar roots, the integrity of the IAC and the visible thinning or fenestration of the lingual cortical wall of the alveolus when the preoperative panoramic radiograph showed darkening of the root.
Materials and methods
This prospective study included 83 patients (39 males and 44 females) with a mean age of 28.1 ± 6.7 years (range 16–48 years). Inclusion criteria were patients with darkening of the root on preoperative radiographs. Darkening was defined as a radiolucent band on the root of the third molar, where the IAC crosses it. Isolated/single darkening was defined as darkening of the root without adjacent panoramic signs. Multiple darkening was defined as darkening of the root and one or more of the adjacent panoramic radiographic signs simultaneously (diversion of the canal, narrowing of the canal, and interruption of the superior cortical line; illustrated in Szalma et al. ).
Patients without darkening on preoperative radiographs were excluded from the study. All the patients underwent surgery in the authors’ department (Department of Oral and Maxillofacial Surgery, University of Pécs, Pécs, Hungary) between January 2010 and September 2011. Prior to operation, each patient provided full informed consent. The study was approved by the Regional Research Ethics Committee of the Medical Center, Pécs (Ref. No. 3795.316-7851/KK4/2010).
Third molar removals were carried out under local anaesthesia. Envelope (sulcular) mucoperiosteal flaps were raised at superficial impactions, and triangular flaps were raised at deep impactions, followed by vestibular and distal bone removal. The sockets were irrigated with 20 ml of sterile saline solution at room temperature, and the alveoli (especially the IAN exposure, root morphology, groove formation and visible lingual cortical defects, perforations or thinning) were checked during and after precisely focused, careful suction. If excessive bleeding made intra-operative observation ambiguous, the case was excluded from the study. The exposed IAN bundles were examined under loupe magnification using a headlight. An IAN exposure was defined upon direct visualization of the suspected neurovascular bundle when the following criteria were partially or totally fulfilled: mesiodistally oriented tubular, pale or whitish structure at the expected level of the socket (estimated according to the panoramic radiographs). The criteria for lingual cortical thinning were a regularly or irregularly shaped dark area on the lingual surface of the alveolus, with or without an impression caused by the third molar. Soft tissues on the lingual surface lacking a tubular shape and lacking mesiodistal orientation were determined to be fenestration of the lingual cortical wall. The root morphology was observed from both mesiodistal and buccolingual directions after removal. If the reconstruction of fractured or segmented roots was not possible, the case was excluded.
When the IAN was visible, iodoform-impregnated drains were placed at the suture insertion to prevent possible nerve compression. Whenever possible, patient documentation included macro-photographs (Canon EOS 500D, Canon EF 100 mm f/2.8 USM macro objective; Canon Macro Ring Lite MR-14EX, Canon, USA).
Digital panoramic radiographs (PaX-400C with a maximum sensor resolution of 10.42 line pairs/mm, Vatech, Korea) were taken before surgery. Preoperative radiographs and the presence of the high risk sign, darkening of the root (with or without adjacent ‘high risk’ signs) were analysed by the first two authors (J.Sz.) and (L.V.). For preoperative panoramic image analysis, the Easydent (Vatech, Korea) software was used. Observations were conducted on a laptop computer (HP Pavilion dv5; 1280 × 800 maximum resolution). Image manipulation using the enhancement tools of the viewer-solution (magnification, contrast, brightness, sharpness, inverted colourization) was permitted. Ambient light during observation was reduced to less than 50 lx. Intra- and inter-observer reliabilities were tested by kappa statistics.
Data collection and statistical analysis were carried out with SPSS 18.0 (SPSS Inc., Chicago, USA) software. Associations of single and multiple darkening cases with the presence of nerve exposure or with the presence of groove/hook formation were tested by Pearson’s χ 2 test. A P value less than 0.05 was considered to be significant. Cohen’s kappa statistic was used to calculate observer agreement. A kappa value of less than 0.40 was considered to show poor agreement, a value of 0.40–0.59 was considered to be fair agreement, a value of 0.60–0.74 was considered to be good agreement, and a value of 0.75–1.00 was considered to be excellent agreement.
Both the intra- and inter-observer reliabilities were excellent in this study. There were no statistical differences between the single and multiple study groups for age and gender ( P = 0.09 and P = 0.13). Overall, 47 patients presented multiple root darkening on panoramic radiographs, and 36 patients showed single darkening. Of the 83 extractions, the IAN was visible in 37 cases (38/83, 45.8%), broken down 30 times (30/83, 36.2%) in multiple darkening cases and 8 times (8/83, 9.6%) in single darkening patients. The localization of the visible IAN was: 10.8% buccal (4/37 cases), 40.6% inferior or inter-radicular (15/37 cases) and 48.6% lingual (18/37 cases). Groove or hook formation was observed 25 times in multiple darkening cases (25/47, 53.2%) and 6 times in single darkening cases (6/36, 16.7%). All grooves were localized on the lingual surfaces of the third molar roots or were positioned inter- or intra-radicularly.
The occurrence and frequency of different aetiological factors of dark band genesis are presented in Fig. 1 . IAN exposure (missing canal wall) with or without groove formation of the root was the most frequent reason for darkening (45.8%), while visible lingual cortical thinning was observed in 26.5% of the cases. In 9.6% of the cases, special root conformation (imitating darkening, Figs 2 and 3 ) was observed without any nerve exposure or cortical deficiency, while 18.1% of the cases had no obvious reason for the genesis of darkening.