The original contribution of this study was to describe and objectify intraoral neurophysiological changes in patients suffering from LNI and patients with an idiopathic BMS. Therefore, we used a standardized QST protocol.
LNI group (A)
Patients with an LNI clinically revealed a profound hypoesthesia. All of them were proved to have a reduced thermal and mechanical sensibility compared to the control group and compared to the contralateral side as well. This is in accordance with Yilmaz et al. [
32] who reported a significant decreased sensitivity to cool temperature and also to warm stimuli on the injured side compared to control side. They also described a cold and heat hypoalgesia on the injured and non-injured side.
The type of injury such as local anesthetic injection, dental treatment, tumor surgery or surgery after recurrence of a peripheral giant cell granuloma did not influence testing results.
Thermal tests included detection and pain thresholds for warmth, hot and cold sensations representing C- and A-delta fiber mediated stimuli [
27,
31]. All thermal tests showed significantly altered thresholds compared to control side and also to healthy individuals as an expression of a sensory deficit of the lingual nerve. According to these results a deficit of these fibers can be assumed. This was also described by Renton et al. [
29], who investigated a reduced sensitivity of the injured side of the tongue to hot and cold stimuli. In their study, sensitivity on cold thresholds proved to be higher than on warm thresholds. This is in line with previous literature (Green 1984 [
33] and 1987 [
34]) and also confirms results of our study in which the cold detection thresholds were increased significantly compared to the contralateral side. This significantly increased threshold for detection of cold stimuli in the affected area indicates a decreased function of slightly myelinated A-delta fibers [
31,
35,
36].
Jääskeläinen et al. [
37] already reported long-lasting thermal hypoesthesia after unilateral nerve injury. They pointed out that thermal hypoesthesia or anesthesia recognized by the patient is characteristic for peripheral nerve injury, as seen in patients of this study.
Small, unmyelinated C-fibers were represented by the WDT and HPT. After an oral surgical treatment, WDT was shown to improve more slowly than the other tests [
16]. This may proof damage to C-fibers as a negative prognostic factor [
16]. Since the thermal thresholds for warm were also increased, it is to assume that damage or violation of unmyelinated C-fibers might also be involved in our patients. As a consequence, peripheral nerve damage was investigated and thermal hypoesthesia [
30] and hypoalgesia could be assumed which is in line with Yilmaz et al. [
32] who also proofed a cold and heat hypoalgesia in a much larger sample size.
In one LNI patient we even found cold hyperalgesia on the uninjured contralateral side as seen in the performed z-score analysis. However, further investigations showed no deviations in all other assessments and were similar to healthy controls.
These findings were in accordance with results of the mechanical tests. Mechanical pain thresholds are more suitable to detect sensory plus and minus signs [
14,
16]. Affected sides of the tongue presented significantly increased PPT thresholds (C- and A-delta fiber mediated), MPT, MDT and MPS (all A-delta fiber mediated stimuli) and VDT compared to contralateral, unaffected side. VDT was significantly increased comparing the affected side of the tongue with the unaffected side, indicating damage in A-beta fibers [
31,
35]. Except for PPT and VDT increased thresholds were also found comparing patients with healthy subjects. Maybe due to small sample size, there was no significant difference comparing the affected side with the volunteers for VDT and PPT. In ANOVA testing, there was no significant difference comparing both sides of the tongue, but with LSD post-hoc test LNI side compared with the volunteers showed a significant difference for WUR. This demonstrated also a deficit of A-beta fibers. This pattern of the mechanical tests showed a deficit of the affected fibers and confirmed the results of the thermal tests.
Using QST we were able to demonstrate a sensory deficit in all fiber functions, painful and non-painful stimuli for peripheral LNI. This result is confirmed by other studies [
10,
30,
38,
39]. As a consequence, it is to assume that patients with LNI suffer from serious axonal injuries or separations with irreversible sensory disturbances. All of them are affected by a primary axonotmesis or neurotmesis, caused by extensive damage or destruction of axons with consecutive decrease in density of all fibers (A-beta-, A-delta-, C-fibers) which explains the sensory deficits.
BMS group (B)
In the BMS group, CDT and WDT also showed significant deficiencies in affected patients indicating a cold/warmth-hypoesthesia and consecutively a small fiber loss. Yilmaz et al. [
32] reported contrary findings. They proofed BMS patients being significantly more sensitive to cold and warm stimuli than healthy controls representing a hyperesthesia. One possible explanation might be a different patient composition which could cause a deviation of results especially in a limited patient cohort. In general, results should be interpreted with consideration of the limitations of the psychophysiological assessment tool which is highly dependant on patients’ mental abilities and compliance. Patients may also get tired and distracted. As reported in this recent study, there is also difficulty in identification of cases of heat hyperalgesia (possibly due to adaptation to warming due to the method of limits (+50 °C cutt-off point). As a possible solution, the Method of Levels was discussed to avoid tired and distracted patients and – concerning ethical guidelines – faster stimulus ramps and suprathresholds.
Evaluating profiles of neuropathic pain states, cold and warm hypoesthesia is well-known [
40] and also found as a part of trigeminal nerve injury [
41]. These peripheral neurologic changes coexist with other neurophysiological findings. For the tongue, a significant cold hyperalgesia was evaluated in CPT pointing to a small fiber neuropathy with central components. This cold hyperalgesia is a sign for central sensitization in patients with BMS, which was not found in patients with LNI. Yilmaz et al. [
32] also documented a cold hyperalgesia compared with control patients.
A possible mechanism for a central sensitization could be found in the plasticity of neurons in the posterior horn of the spinal cord. Another explanation may be a loss of the central pain inhibition normally performed by the brain stem, the rostral ventro-medial medulla and the central gray matter. These results are also in accordance with Lauria et al. [
42], who analyzed the tongues of BMS-patients by histological examination and according to Forssell et al. [
30]. Both studies concluded that BMS has a neuropathic component. Primary burning mouth syndrome was also emphasized to be caused by deregulated subclinical neuropathic pain [
43].
Mechanical tests (especially MPT) revealed a pinprick-hypoalgesia indicating an impaired function in small fibers. As already seen in our LNI patients, this hypofunction is a typical finding in patients suffering from peripheral nerve damage. This is in accordance with previously published results [
37].
Comparing measurements of the dorsal foot in affected patients and healthy control group, all QST parameters revealed no significant differences. By using the dorsal foot as reference, a generalized neuropathy could be excluded. In this study, BMS was evaluated as a localized neuropathy with variable central and peripheral contributions among individuals as already described elsewhere [
43].
Profiling patients with LNI and BMS, QST proved to be a non-invasive, psychophysical approach to profile thermal and mechanical somatosensation as already shown in previous studies [
16,
27,
30,
44,
45]. Sensory signs were characterized by QST. They point to possible neurobiological mechanisms such as central or peripheral sensitization (Yekta et al., 2010b, Rolke et al., 2006a, Rolke et al., 2006b, Said-Yekta et al., 2012, Cruccu et al., 2004, Renton et al., 2006).
Because of increasing forensic implication in patients’ treatment, QST might be a useful tool to objectify clinical findings [
14]. In affected patients with LNI or BMS, the monitoring of afferent nerve fiber functions is challenging and in these cases, QST might help to support decisions on further interventions. This was already emphasized by Jaaskelainen et al. [
43] who described three distinct subclinical neuropathic pain states in BMS patients. They emphasized targeted treatment modalities according to these subgroups and careful neurophysiologic examinations to distinguish these groups. Also in patients with iatrogenic nerve lesions, targeted treatment modalities were presented in previous studies [
9,
46]. A proper use of neurophysiological diagnostic tests seems to offer the possibility for targeted interventions based one the underlying pathophysiological mechanisms.
As an alternative to QST, sensory dysfunction in humans can be objectively quantified by electrophysiological recordings of trigeminal sensory-evoked cortical potentials [
47] and brainstem reflexes [
48] after stimulation of extraoral and intraoral sites [
44,
49]. Functional magnetic resonance imaging may also reveal sensory functions [
50,
51]. But these methods are complex, and still more time consuming than QST and do not allow the identification of isolated sensory deficiencies and therefore seem not to be appropriate in clinical routine. The QST protocol used in this study remains time consuming in clinical routine diagnostic. Some studies [
29,
52] restricted screening to thermal QST to examine the tongue. By extending the protocol to include mechanical components however, a differentiated classification of nerve fiber function is possible. Previous studies [
6,
53] emphasized to reduce the QST battery to only seven parameters. In order to benefit from the entire potential of QST based diagnosis, exclusion of parameters can’t be recommended because essential information concerning hypo- and hyperesthesia would be missing.