Functional Imaging in Olfactory Disorders

Purpose of Review

The aim was to synthesize key findings regarding the use of functional MRI (fMRI) to assess olfactory dysfunction (OD), and thus, to evaluate whether fMRI could be a reliable clinical diagnostic tool.

Recent Findings

In response to olfactory stimulation, patients with quantitative OD display reduced activation in olfactory-related brain regions but also stronger activation in non-olfactory brain areas. Parosmic patients also seem to show both weaker and higher brain signals. As to trigeminal chemosensory system, fMRI suggests that central processing may be declined in patients with OD. Functional connectivity studies report a possible correlation between altered neuronal connections within brain networks and olfactory performances.

Summary

fMRI emerges as a valuable and promising objective method in OD evaluation. Yet, its high inter-individual variability still precludes its routine clinical use for diagnostic purpose. Future research should focus on optimizing stimulation paradigms and analysis methods.

Neural correlates of olfactory processing in congenital blindness

Adaptive neuroplastic changes have been well documented in congenitally blind individuals for the processing of tactile and auditory information. By contrast, very few studies have investigated olfactory processing in the absence of vision. There is ample evidence that the olfactory system is highly plastic and that blind individuals rely more on their sense of smell than the sighted do. The olfactory system in the blind is therefore likely to be susceptible to cross-modal changes similar to those observed for the tactile and auditory modalities. To test this hypothesis, we used functional magnetic resonance imaging to measure changes in the blood-oxygenation level-dependent signal in congenitally blind and blindfolded sighted control subjects during a simple odor detection task. We found several group differences in task-related activations. Compared to sighted controls, congenitally blind subjects more strongly activated primary (right amygdala) and secondary (right orbitofrontal cortex and bilateral hippocampus) olfactory areas. In addition, widespread task-related activations were found throughout the whole extent of the occipital cortex in blind but not in sighted participants. The stronger recruitment of the occipital cortex during odor detection demonstrates a preferential access of olfactory stimuli to this area when vision is lacking from birth. This finding expands current knowledge about the supramodal function of the visually deprived occipital cortex in congenital blindness, linking it also to olfactory processing in addition to tactile and auditory processing.

TMS of the occipital cortex induces tactile sensations in the fingers of blind Braille readers

Various non-visual inputs produce cross-modal responses in the visual cortex of early blind subjects. In order to determine the qualitative experience associated with these occipital activations, we systematically stimulated the entire occipital cortex using single pulse transcranial magnetic stimulation (TMS) in early blind subjects and in blindfolded seeing controls. Whereas blindfolded seeing controls reported only phosphenes following occipital cortex stimulation, some of the blind subjects reported tactile sensations in the fingers that were somatotopically organized onto the visual cortex. The number of cortical sites inducing tactile sensations appeared to be related to the number of hours of Braille reading per day, Braille reading speed and dexterity. These data, taken in conjunction with previous anatomical, behavioural and functional imaging results, suggest the presence of a polysynaptic cortical pathway between the somatosensory cortex and the visual cortex in early blind subjects. These results also add new evidence that the activity of the occipital lobe in the blind takes its qualitative expression from the character of its new input source, therefore supporting the cortical deference hypothesis.

rTMS of the occipital cortex abolishes Braille reading and repetition priming in blind subjects

To study the functional involvement of the visual cortex in Braille reading, we applied repetitive transcranial magnetic stimulation (rTMS) over midoccipital (MOC) and primary somatosensory (SI) cortex in blind subjects. After rTMS of MOC, but not SI, subjects made significantly more errors and showed an abolishment of the improvement in reading speed following repetitive presentation of the same word list, suggesting a role of the visual cortex in repetition priming in the blind.

Stereoscopic processing in the human brain as a function of binocular luminance rivalry

We investigated the neural substrates of a recent model of human stereodepth perception by obtaining measurements of regional cerebral blood flow (rCBF) using PET. Subjects experienced the perceptual properties of stereopsis by viewing rival-luminance stereograms displaying an identical random-dot pattern in their central portion while the backgrounds exhibited correspondent dots contrasting in black/white luminance. The stereoscopic vision induced by retinal luminance rivalry coincided with a significant elevation of rCBF in the dorsal visual pathway. Area V5 (MT) was activated bilaterally by the experimental condition while the remaining active loci were restricted to the right hemisphere. The neural sites that responded to this novel stereoscopic stimulus are similar to those activated by traditional stereograms containing horizontal disparities.

Cortical representation of inward and outward radial motion in man

We have used positron emission tomography to investigate the cortical areas of the normal human brain involved in processing inward (Expansion) and outward (Contraction) radial motion simulated with an optic flow stimulus. The optical flow display was made out of dots moving radially away from or toward the center of the display monitor. In the Control condition, the dots' motion was randomized in order to remove any sensation of radial motion. In the Expansion condition, several loci of activation were observed: visual areas V2-V3 and the superior parietal lobule (BA 7), predominantly in the right hemisphere. In the Contraction condition activation sites were found in the same visual areas (V2 and V3) in the right hemisphere but the increase in rCBF in these regions was much lower than in the Expansion condition. BA 7 was activated in both hemispheres. When the motion component of the stimulus was isolated by subtracting the static condition from the incoherent motion condition, we obtained activations of areas V2, V3, and MT (putative V5). These results indicate that the detection of radial motion derived from an optic flow stimulus is mediated by structures forming the dorsal part of the visual cortical system and confirm that area MT is not specifically involved in flow analysis.

Facial pain: from animal models to functional neuroimaging studies

This paper summarizes some recent findings on the physiopathology of facial pain. Over the past decade, a number of animal models of facial pain have been developed. Two of these models are discussed in more detail. The model developed by Strassman and Burstein has provided a useful tool for the study of the mechanisms underlying migraine. Single unit recordings revealed that medullary dorsal horn neurons show a long-lasting increased response to dural and cutaneous periorbital mechanical and thermal stimulation after application of inflammatory agents to the dura. In addition, dural and cutaneous receptive fields largely expanded and spontaneous ongoing activity developed. These findings suggest that the extracranial hypersensitivity that is often observed in headache may have an intracranial origin. The second model that is discussed is the model of facial neuropathic pain after infraorbital nerve ligation developed by Vos and colleagues. In contrast to the previous model which is an acute electrophysiological preparation, the infraorbital nerve model is a behavioural model. It can therefore be used both for electrophysiological and behavioural studies. In recent years, a number of functional neuroimaging studies of facial pain appeared. Studies performed in cluster headache patients seem to point to a crucial role of the hypothalamus in this syndrome. However, since hypothalamic activation has also been reported in some other types of (facial) pain, the specificity of the role of the hypothalamus in cluster headache remains to be proven.

Photochemically-induced ischemia of the rat sciatic nerve produces a dose-dependent and highly reproducible mechanical, heat and cold allodynia, and signs of spontaneous pain

Sensory abnormalities and changes in spontaneous behavior were examined after a photochemically induced ischemic lesion of the rat sciatic nerve. Male adult rats were anesthetized and the sciatic nerve was exposed. After the intravenous injection of a photosensitizing dye, erythrosin B, the exposed nerve was irradiated just proximal to the nerve trifurcation with light from an argon laser. Three different irradiation times were used, 30 s, 1 and 2 min. In sham-operated rats, the exposed sciatic nerve was irradiated for 2 min without prior injection of the erythrosin B. Rats were tested for the presence of mechanical, cold and heat allodynia or hyperalgesia. All the animals in the 1- and 2-min irradiation groups developed mechanical, cold and heat allodynia after nerve irradiation. A significant dose-dependent effect of laser exposure time was observed for all modalities tested (2 min > 1 min > 30 s = sham). The maximum effects were observed at 3 and 7 days postirradiation and remained present for up to 10 weeks. No significant contralateral effects were observed in any of the groups. In three separate groups of rats (1, 2 and 4 min of laser exposure), the presence of possible signs of spontaneous pain (paw shaking, paw elevation and freezing behavior) was tested. A significant and exposure time-dependent increase in spontaneous paw elevation and paw shaking was observed which was maximal at week 1, but resolved at 4 weeks (4 min > 2 min > 1 min > sham). In addition, animals in all ischemic groups, but not in the sham group, showed a significant increase in freezing behavior up to 4 weeks after nerve irradiation. Light microscopic evaluation of nerves removed 7 days post-irradiation, i.e. when maximal allodynia was observed, showed clear evidence of demyelination of large myelinated fibers. These data indicate that photochemically-induced peripheral nerve ischemia is associated with abnormal pain-related behaviors, including mechanical, thermal and cold allodynia and signs of spontaneous pain. The incidence and severity of the behavioral changes are clearly dependent on the exposure time and are probably due to, at least in part, a demyelinaton. These results partly confirm previous data using a similar technique and suggest that this may represent a new animal model for peripheral neuropathy of ischemic origin. The advantages of the present model are its good reproducibility and the fact that the nerve injury can be easily quantified and graded.

Systemic lidocaine induces expansion of the receptive field of spinal dorsal horn neurons in rats

The effect of intravenous administration of the local anaesthetic lidocaine (1, 3 or 5 mg/kg) on the responsiveness and size of the cutaneous receptive fields of 18 lumbar dorsal horn neurons was examined in intact urethane-anaesthetized rats. Lidocaine induced expansion of the receptive field in the majority of neurons examined, particularly after the two higher doses. The expansion occurred usually within 10 min after lidocaine injection and the effect was reversible. Lidocaine also altered the responsiveness of dorsal horn neurons to peripheral mechanical stimulation. The responses of wide-dynamic-range neurons to noxious pinch were usually inhibited by lidocaine. However, some low-threshold neurons started to react to noxious mechanical stimulation and some high-threshold neurons started to respond to innocuous brushing after lidocaine injection. The present results show that moderate doses of systemic lidocaine induce complex changes in the excitability of dorsal horn neurons, including an increase in the size of the receptive field and altered response characteristics to mechanical stimulation.

The consumption of fentanyl is increased in rats with nociceptive but not with neuropathic pain

The question whether opioids relieve neuropathic pain remains a controversial issue. Experimental as well as clinical studies report contradictory results. This study investigated the consumption of fentanyl, a short-acting opioid, in rats with neuropathic pain, induced by partial sciatic nerve injury. The experiment consisted of a drug choice procedure in which the animals could choose between a solution containing 0.008 mg/ml of fentanyl and a highly palatable sweet solution. It was reasoned that if opioids have an analgesic effect in neuropathic pain, this will reinforce the intake of fentanyl more so in rats with neuropathic pain than in pain-free controls. This protocol was previously already used by Colpaert et al. (1982) in a rat model of chronic pain of nociceptive origin, namely polyarthritis. No significant differences were found in the relative oral intake of the fentanyl solution in mononeuropathic and pain-free control rats. In contrast, rats with nociceptive pain, adjuvant monoarthritis, drank significantly more of the fentanyl solution than did control rats. These data give experimental support for the clinical findings that opioids have a poor analgesic effect in neuropathic pain.

Therapeutic stereotactic procedures on the thalamus for pain

Thalamotomy and electrical stimulation of a thalamic target as treatment for persistent pain are discussed. Thalamotomy is only rarely performed these days according to a questionnaire, given to some colleagues, about the type and the number of operations they performed in the years 1984, 1985 and 1986. The need for stimulation in the periventricular or periaqueductal grey for nociceptive pain is decreasing due to the advent of intraspinal and intraventricular administration of opioids. Nowadays medial and lateral ventro-posterior thalamic nuclei are frequently stimulated for treatment of deafferentation pain. Of 36 patients with deafferentation pain, 22 initially had benefit from this stimulation, but long-term success was only achieved in 11 (30%) of them. It was a general trend that the patients with an initial high pain relief score obtained the best long-term results.

What can the neurosurgeon offer in peripheral neuropathic pain?

Neurosurgery has much to offer in the treatment of peripheral neuropathic pain but selection of the best procedure for a given patient remains problematic: planning of the treatment must be based on an analysis of the pathophysiological mechanism in the given case but the identification of this mechanism is often difficult. Available procedures are: 1) Nerve repair, neurolysis and nerve relocation; 2) Interventions on the sympathetic nervous system; 3) Neurostimulation; 4) Intraspinal morphine; 5) Ablative lesions. Neurosurgeons have, or should have, the necessary neuroscience background and microsurgical skills to be important partners of the team caring for patients with peripheral neuropathic pain.

Further evidence for myelinated as well as unmyelinated fibre damage in a rat model of neuropathic pain

A mononeuropathy, produced by ligation of the sciatic nerve in rats, has recently been proposed as an animal model of experimental pain and pain-related disorders (hyperalgesia and allodynia). We investigated quantitatively the morphological changes in myelinated and unmyelinated fibres of the sciatic nerves 2 weeks after ligation in rats exhibiting allodynia to thermal stimulation. There was a marked reduction in the number of large myelinated fibres distal to the ligature (711 +/- 34 compared with 5315 +/- 230 in normal nerves). We also found a significant loss of small myelinated fibres (2429 +/- 109 compared with 3197 +/- 308 in normal nerves), the remaining fibres of this type showing pathological properties. Finally, ultrastructural evidence of damage to unmyelinated fibres was found. The typical pattern of large clusters of normal unmyelinated axons was no longer present within most regions of the nerve. There was a significant reduction in the size of the unmyelinated fibres (0.41 micron +/- 0.15 compared with 0.71 micron +/- 0.08 in normal nerves), together with a twofold increase in their number per cluster. Hypotheses about the mechanism of thermal allodynia in this pain model therefore must take into account the fact that all fibre classes show pathological changes.

Deep brain stimulation in the treatment of chronic pain in man: where and why?

In current clinical practice, two brain structures are stimulated for the relief of chronic pain, namely the somatosensory thalamic nuclei (VPL-VPM) and the periventricular and periaqueductal gray matter (PVG-PAG). Whereas stimulation of the VPL-VPM is almost exclusively used for the treatment of deafferentation pain, stimulation of the PVG-PAG is mostly used in cases of nociceptive pain. We present our results of VPL-VPM stimulation in 36 patients with deafferentation pain. Initial pain relief was obtained in 61% of patients. To-day, after a mean follow-up of more than 4 years, 30% are still pain free. This success rate was found to be lower than the mean reported success rate of 57%, based on a survey of the world literature. Upon reviewing the literature, it was apparent that the reported success rates vary considerably between different authors. Some tentative explanations are given for this large discrepancy in success rate. The mechanisms by which electrical stimulation of the VPL-VPM suppresses deafferentation pain remain to be elucidated. Recent clinical and experimental findings suggest that a dopaminergic mechanism might be involved.