Neurosensory disturbance after sagittal split osteotomy is a common complication. This study evaluated the course of the mandibular canal at three positions using computed tomography (CT), assessed the risk of injury to the inferior alveolar nerve in classical sagittal split osteotomy, based on the proximity of the mandibular canal to the external cortical bone, and proposed alternative surgical techniques using computer-assisted surgery. CT data from 102 mandibular rami were evaluated. At each position, the distance between the mandibular canal and the inner surface of the cortical bone was measured; if less than 1 mm or if the canal contacted the external cortical bone it was registered as a possible neurosensory compromising proximity. The course of each mandibular canal was allocated to a neurosensory risk or a non-neurosensory risk group. The mandibular canal was in contact with, or within 1 mm of, the lingual cortex in most positions along its course. Neurosensory compromising proximity of the mandibular canal was observed in about 60% of sagittal split ramus osteotomy sites examined. For this group, modified classic osteotomy or complete individualized osteotomy is proposed, depending on the position at which the mandibular canal was at risk; they may be accomplished with computer-assisted navigation.
Sagittal split ramus osteotomy (SSRO) of the mandibular ramus was reported to have been introduced in 1942 . The credit for improving this osteotomy goes to T rauner & O bwegeser who, in 1957, described their modified sagittal split osteotomy. Since 1957, SSRO has become a standard procedure in the treatment of mandibular deformity . SSRO has become a versatile technique to advance and set back the mandible . General acceptability of this technique by maxillofacial surgeons has led to various modifications by many clinicians . These modifications have made the technique easier and more predictable .
Despite its versatility and the numerous advantages of SSRO, neurosensory disturbances after the operation are common , because it is performed in close proximity to the inferior alveolar nerve (IAN) . Neurosensory disturbance is reported to develop in the lower lip and mental skin of 30–40% of patients after such surgery . Factors that influence neurosensory disturbance after SSRO include: age of patient, intraoperative magnitude of mandibular movement, degree of manipulation of the IAN and the width of marrow space between the mandibular canal and the external cortical bone .
Y amamoto et al. showed that neurosensory disturbance was significantly more likely to be present 1 year after surgery when the width of the marrow space between the mandibular canal and the external cortical bone was 0.8 mm or less. It has been reported that separating the IAN from the external cortical bone without injuring the IAN canal is difficult when a marrow space is absent . It has been suggested that the width of the marrow should be considered when planning the treatment of patients undergoing SSRO, and in some cases, the surgeon should select a procedure other than SSRO to avoid nerve injury .
Computer-assisted surgery technology has been employed in several surgical fields such as neurosurgery, endoscopy, arthroscopy, and bone surgery . Reducing the risk of damage to anatomical structures such as nerves, vessels and neighbouring structures is one of the desired outcomes of preoperative computer-aided planning . Tools for surgical guidance aim to transfer preoperative planning based on volumetric patient data (computed tomography (CT) or cone-beam CT (CBCT)) to the intra-operative site . Computer-assisted navigation allows for real-time imaging of the surgical drill as an overlay graphic on CT and live intra-operative video images and has been reported to be suitable for routine clinical applications .
The aim of this study is to evaluate the course of the mandibular canal using CT, to assess the risk of injury to the IAN in classical sagittal split osteotomy based on the proximity of the mandibular canal to the external cortical bone, and to propose alternative surgical techniques using computer-assisted intra-operative navigation.
Materials and methods
CT data (right mandible:left mandible = 52:50) preoperatively acquired for navigated bimaxillary osteotomy in 52 patients (31 male, 21 female) were retrospectively evaluated to determine the course of the mandibular canal. At each CT slice (positions X1, X2, X3) the distance between the mandibular canal and the inner surface of the cortical bone was measured and registered as a possible neurosensory compromising proximity if it was less than 1 mm or if the mandibular canal came into contact with the external cortical bone ( Fig. 1 ). At position X3 the location (upper, middle, lower third) of the canal was investigated ( Fig. 1 ).
Images were acquired using different CT scans. Bone reconstruction mode was used. The data acquisition protocol was optimized for navigation purposes, with a gantry tilt of 0°. The CT scans were taken parallel to the Frankfort plane at 0.5 mm intervals, with a slice thickness of 0.5 mm. Voxel size was 0.17 mm × 0.17 mm × 0.75 mm on a 1024 × 1024 matrix, with a field of view of 175 mm × 175 mm. Volume images were transferred to the Surgicase CMF ® software (Materialise N.V., Leuven, Belgium) using the standard DICOM protocol.
The following definitions were used. Neurosensory compromising proximity was recorded if the distance between the mandibular canal and the inner surface of the cortical bone was less than 1 mm or there was contact between the mandibular canal and the inner buccal cortical bone. Proximity at the lingual cortex was not considered a neurosensory compromising proximity. Thin mandible (Tn) was recorded for proximity at both the buccal and lingual cortices (neurosensory compromising proximity) ( Fig. 2 a–c ). Thick mandible (Tk) was recorded for no proximity at both the buccal and lingual cortices (no neurosensory compromising proximity) ( Fig. 3 a–c ). Buccal proximity (Bn) was recorded for proximity only at the buccal cortex with nerve at risk of injury (neurosensory compromising proximity) ( Fig. 4 a–c ). Lingual proximity (Ln) was recorded for proximity only at the lingual cortex (no neurosensory compromising proximity) ( Fig. 5 a–c ). Based on the above, only Tn and Bn were considered as neurosensory compromising proximities.