Iatrogenic damage to the mandibular nerves as assessed by the masseter inhibitory reflex

Our study in a large cohort of patients now shows the diagnostic accuracy of the MIR, a standard neurophysiological tool, in demonstrating iatrogenic damage to the mandibular nerves. We found that the MIR had 99% specificity and 51% sensitivity, i.e. MIR testing reliably demonstrates nerve damage beyond doubt, whereas the finding of a normal MIR is by no means sufficient to exclude nerve damage. Probably the dysfunction of a small number of nerve fibres, insufficient to produce a MIR abnormality, may still engender important sensory disturbances.

We propose that MIR testing, when used for legal purposes, be considered reliable in one direction only, i.e. abnormality does prove nerve damage, normality does not disprove it.

Although it would be reasonable to expect some degree of recovery with time, because of spontaneous healing and reinnervation, we found no correlation between the delay since injury and MIR abnormality. This is probably due to the characteristics of our patient sample: all patients were examined when at least 2 months had elapsed from injury (i.e. when no further healing is expected to occur) and 146 out of 160 were examined beyond 4 months (i.e. when, given the short length of the lingual and inferior dental nerves, the process of reinnervation was completed). Given the importance of this aspect, however, we are now planning a study in the acute phase after injury.

Other neurophysiological tests may also assess mandibular nerve function: trigeminal-evoked potentials, nerve conduction study, and blink reflex. Although some authors used trigeminal somatosensory-evoked potentials to study IAN injuries after surgical procedures, the reliability of these signals has been questioned because, rather than reflecting genuine brain activity, probably they result from volume-conducted muscle signals, as they disappear in the curarized subject [10, 11]. Consistently, the IFCN recently recommended to investigate the trigeminal function with reflex rather than evoked potential studies [12]. The IAN nerve conduction study is a reliable method, but it is undeniably invasive (the recording needle-electrode is inserted below the zygomatic arc to a depth of about 4.5 cm) [13]. Unlike the above techniques, the blink reflex after mental or lingual stimulation seems a promising alternative to the MIR. It has been widely studied in patients with orofacial pains or trigeminal neuropathy and, in combination with the IAN conduction study, showed sensitivity (59%) and specificity (100%) values very similar to those we found in our patients. The patients with iatrogenic damage after dental procedures, however, still represent a comparatively small sample [13, 14]. Damage to the lingual nerve after third molar extraction can also be tested with an interesting method, very similar to ours (even though the main measure is not the reflex latency) [15]. Also in this case, the number of patients is still too small to test diagnostic accuracy.

Whereas most patients had hypoesthesia, less than one-third had pain, probably because mechanical injury mainly damages large myelinated, non-nociceptive fibres, and tends to spare small nociceptive fibres. Accordingly, because MIR is mediated by non-nociceptive Aβ fibres [4], and it does not provide any information on nociceptive pathways, we found that MIR abnormalities were unrelated to pain. Alternatively, the complexity of pain mechanisms at orofacial level simply makes it unreasonable to expect a direct correlation between pain and number of damaged primary afferents [16]. In particular, we cannot exclude that plastic changes in the central nervous system contributed to the sensory disturbances reported by our patients. According to this hypothesis, sensory disturbances could be triggered by the nerve damage and persist after nerve recovery due to central mechanisms such as central sensitization.