The objective of this study was to improve the accuracy of diagnosis of inferior alveolar nerve (IAN) injury by determining degrees of nerve disturbance using the sensory nerve action potential (SNAP) and sensory nerve conduction velocity (SCV). Crush and partial and complete nerve amputation injuries were applied to the IAN of rabbits, then SNAPs and histomorphometric observations were recorded at 1, 5, and 10 weeks. For crush injury, most nerves were smaller in diameter at 5 weeks than at 1 week, however after 10 weeks, extensive nerve regeneration was observed. The SNAP showed a decrease in SCV at weeks 1 and 5, followed by an increase at week 10. For partial nerve amputation, small to medium-sized nerve fibres were observed at weeks 1 and 5, then larger nerves were seen at week 10. Minimal changes in SCV were observed at weeks 1 and 5, however SCV increased at week 10. For complete nerve amputation, nerve fibres were sparse at week 1, but gradual nerve regeneration was observed at weeks 5 and 10. SNAPs were detectable from week 10, however the SCV was extremely low. This study showed SCV to be an effective factor in the evaluation of nerve injury and regeneration.
Inferior alveolar nerve (IAN) injury is common in the field of oral surgery. IAN injuries are caused not only by complications during mandibular third molar extraction, cyst removal, tumour resection, and any orthodontic surgical procedure, but can also be caused by mandibular osteomyelitis and may even occur during root canal treatment of the mandibular molars. More recently, IAN injury has also been reported to occur during implant placement surgery.
In order to understand the symptoms associated with IAN injury, subjective symptoms are recorded and sensory tests are used. In addition to the two-point discrimination test and pain and temperature tests, the Semmes–Weinstein (SW) pressure aesthesiometer is also used to perform a tactile threshold test (SW test). Unfortunately, the results of these evaluations are determined by patient self-report and their subjective nature makes the accurate diagnosis of nerve injury very difficult. There have been cases where procedures such as nerve grafting or nerve repair surgery have not been undertaken due to the uncertain outcome of the evaluation. In other more severe cases in which follow-up treatment instead of a surgical procedure has been done after subjective evaluation, dysesthesia has occurred as a neurological sequela.
The electrophysiological evaluation of nerve function using the sensory nerve action potential (SNAP) is important for the diagnosis and treatment of nerve injuries, because this evaluation includes the sensory nerve conduction velocity (SCV), amplitude, and waveforms of nerves, allowing the presence and degree of nerve damage, or the progress of nerve repair, to be determined in an accurate and objective manner. Little research has been done on the clinical condition of the IAN injury in relation to SNAP waveforms. Most studies involving these methods have been performed on limbs. With regard to the IAN, Yamazaki and Noma were the first to study IAN damage with compound electrometry; this study was performed on rabbits. Matsuda studied IAN injury in human volunteers, and Sasaki et al. studied IAN complications as a consequence of jaw surgery in humans. Nevertheless, the information reported by these studies is insufficient to provide a fundamental understanding of IAN injury.
The objective of the present study was to improve the accuracy of nerve injury diagnosis by determining the degree of damage using the SNAP and SCV.
Materials and methods
Adult male rabbits weighing 2.5 kg ( n = 32) were prepared. The rabbits were maintained on a diet of pellets and tap water, and those with no abnormal symptoms after a week of care were used for the experiment. All procedures were carried out in accordance with the guidelines for the treatment of experimental animals of Tokyo Dental College. The rabbits were administered with 1% pentobarbital sodium intravenously (0.5 mg/kg) and anaesthetized. The mandibular area was injected with 2% lidocaine (1/80,000 adrenaline) as a local anaesthetic, then a skin incision was made at the inferior border of the mandible and this region was dissected. The inferior border of the mandible and the cortical bone (10 mm × 2 mm) were removed using dental burs and a bone chisel, exposing the IAN. There were three experimental groups of animals. In the crush injury group, the IAN was removed carefully from the IAN canal and then the nerve was crushed with Pean forceps at 750 g force. In the partial nerve amputation group, the IAN was cut half way with a scalpel. In the complete nerve amputation group, the IAN was cut completely with a scalpel. All injuries were applied over a length of 3 mm ( Fig. 1 ). After applying the injury, the cortical bone was placed back in position and stabilized with periosteum sutures, and the surgical incision was closed. The rabbits were then returned to the care facility to recover until the next surgical session. Control group animals were not subjected to any IAN injury.
After 1, 5, or 10 weeks, the rabbits were anaesthetized with the intravenous administration of 1% pentobarbital sodium (0.5 mg/kg). A skin incision was made along the inferior border of the mandible and the mental foramen was exposed, then a nerve-stimulating electrode (TUN209-012A; Unique Medical Co., Ltd., Tokyo, Japan) was placed near the mental foramen. The medial pterygoid muscle was separated from the mandibular angle until the mandibular foramen was exposed, and a recording electrode (NM-330T; Nihon Koden, Tokyo, Japan) was placed near the mandibular foramen ( Fig. 2 ). The control group animals were also anaesthetized and tested in the same manner ( n = 3).
The recording electrode used was a coaxial needle electrode. The SNAP was acquired using an electromyogram and evoked-potential testing device (MEB-9404 Neuropack S1; Nihon Koden). Constant current stimulations of square waves were applied at supramaximal stimulus for durations of 0.1 ms and at a frequency of 1 Hz. The information was recorded by orthodromic method and averaged over 30 trials, and the SNAP was derived. The location of the SNAP measurement was selected with reference to the study of Matsuda. The SCV was calculated using the following equation: SCV (m/s) = latency (ms)/distance (mm). The distance between the mental foramen and the mandibular foramen was used as the transmission distance.
After derivation of the SNAP, the rabbits were euthanized by intravenous injection of 1% pentobarbital. The mandible from the mental foramen region to the mandibular angle was removed and subjected to histomorphometric analysis. The extracted specimens were fixed in paraffin and demineralized, and transverse and longitudinal sections were acquired. Longitudinal sections were made from the central portion of the injured nerves. Transverse sections were taken from a region at least 5 mm from the damaged centre at the distal end ( Fig. 3 ). The transverse sections underwent Klüver–Barrera staining (double-stained with luxol fast blue and cresyl violet), and for each of the sections, nerve bundles were split into 10 fields of 50 μm × 50 μm. The diameters of the nerves in each field were measured and recorded, and histograms were made of the nerve diameter distribution. A total of three sections from three different specimens were examined, resulting in a total of 30 observed fields.
For the statistical analysis, comparisons of the SCV were performed using one-way factorial analysis of variance (ANOVA) and multiple comparison tests. Statistical significance was established at P < 0.05.
Figure 4 summarizes the periodical changes in SNAP for each group. Table 1 shows the actual SNAP for each injury at 1 week, 5 weeks, and 10 weeks. The control group ( n = 3) displayed a diphasic pattern, consisting of a short latency period and a high amplitude wave ( Fig. 4 ).
|Amplitude (mA) a||Latency (ms)||Distance (mm) b||SCV (m/s) c|
|Partial nerve amputation|
|Complete nerve amputation|
In the crush injury group at 1 week after injury, the latency period was extended, SCV was delayed, and the waves had multimodal low amplitudes ( n = 3). At 5 weeks, the latency period was greater than at 1 week, and SCV was delayed. The amplitudes of waves were slightly smaller ( n = 3). At 10 weeks, the latency period was significantly shorter and SCV was greater than at 5 weeks, and these results were statistically significant ( n = 3) ( Figs 4 and 5 ).
In the partial nerve amputation group at 1 week after injury, the latency period was extended, SCV was delayed, and the waves had multimodal low amplitudes ( n = 3). Five weeks following injury, the latency period was slightly extended, SCV was slightly decreased, and wave amplitudes also decreased ( n = 3). At 10 weeks, the latency period was shorter and the SCV greater compared to week 5. In addition, wave amplitudes were greater and almost equal to the values at week 1 ( n = 3) ( Figs 4 and 5 ).
In the complete nerve amputation group, the SNAP could not be derived at 1 week after injury ( n = 3). At 5 weeks, low electric potential values were seen in some rabbits; however, the latency period was extremely long and the SCV was greatly delayed ( n = 4). At 10 weeks, the SNAP was expressed in most rabbits and SCV was slightly higher; however, the amplitudes remained low as in previous weeks ( n = 4) ( Figs 4 and 5 ).
In the control group, all regions of the longitudinal sections displayed a strong reaction to S100 protein, and the nerve fibres were straight and evenly distributed ( Fig. 6 ). Transverse sections showed thick layers of myelin sheath and clearly visible wheel-axle structures. These nerve fibres were well stained and found as bundles in close arrangement ( Fig. 7 ). Assessing the predominant nerve fibres in the control group, healthy nerve fibres were found to be approximately 3–5 μm in diameter. Thus nerves of 3–5 μm in diameter were considered medium-sized nerves, nerves of less than 3 μm in diameter as small nerves, and nerves greater than 5 μm as large nerves ( Fig. 8 ).