Académique Documents
Professionnel Documents
Culture Documents
Dolor Neurofisio
Dolor Neurofisio
www.sciencedirect.com
REVIEW/MISE AU POINT
L. Garcia-Larrea ∗
Inserm U1028, Central Integration of Pain Unit, Centre for Neuroscience of Lyon, University Claude-Bernard Lyon, University
Hospital Pain Center (CETD), Neurological Hospital, 69003 Lyon, France
KEYWORDS Summary Neurophysiological techniques help in diagnosis, prognosis and treatment of chronic
Pain; pain, and are particularly useful to determine its neuropathic origin. According to current stan-
Chronic pain; dards, the diagnosis of definite neuropathic pain (NP) needs objective confirmation of a lesion
Neuropathic pain; or disease of somatosensory systems, which can be provided by neurophysiological testing.
Evoked potentials; Lesions causing NP mostly concern the pain-temperature pathways, and therefore neurophysi-
Diagnosis; ological procedures allowing the specific testing of these pathways (i.e., A-delta and C-fibres,
Therapy; spino-thalamo-cortical tracts) are essential for objective diagnosis. Different techniques to
Human stimulate selectively pain-temperature pathways are discussed. Of these, laser-evoked poten-
tials (LEPs) appear as the easiest and most reliable neurophysiological method of assessing
nociceptive function, and their coupling with autonomic responses (e.g., galvanic skin response)
and psychophysics (quantitative sensory testing - QST) can still enhance their diagnostic yield.
Neurophysiological techniques not exploring specifically nociception, such as standard nerve
conduction velocities (NCV) and SEPs to non-noxious stimulation, should be associated to the
exploration of nociceptive systems, not only because both may be simultaneously affected
to different degrees, but also because some specific painful symptoms, such as paroxysmal
discharges, may depend on specific alteration of highly myelinated A-beta fibres. The choice
of techniques is determined after anamnesis and clinical exam, and tries to answer a num-
ber of questions: (a) is the pain-related to injury of somatosensory pathways?; (b) to what
extent are different subsystems affected?; (c) are mechanisms and lesion site in accordance
with imaging data?; (d) are results of use for diagnostic or therapeutic follow-up? Neuropathic
pain (NP) affects more than 15 million people in Western countries, and its belated diagnosis
leads to insufficient or delayed therapy. The use of neurofunctional approaches to obtain a
‘‘physiological photograph’’ of somatosensory function is therefore highly relevant, as it yields
significant clues about the type and mechanisms of pain, thus prompting rapid and optimised
therapy.
© 2012 Elsevier Masson SAS. All rights reserved.
∗ Hôpital neurologique, 59, boulevard Pinel, 69003 Lyon, France. Tel.: +33 4 72 35 78 88.
E-mail address: luis.larrea@univ-lyon1.fr
0987-7053/$ – see front matter © 2012 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.neucli.2012.03.001
188 L. Garcia-Larrea
Introduction and general considerations As the term ‘‘dysfunction’’ is imprecise and a source of mis-
understanding, a proposal of change was launched by the
This article aims to provide a concise approach to the neu- Special Interest Group on NP from the International Asso-
rophysiological assessment of the chronic pain patient, and ciation for the Study of Pain (IASP), which suggested to
addresses the questions ‘‘why, when and how to do it’’. restrict the definition of NP to pain ‘‘arising as a direct
Electrophysiological approaches give access to the study consequence of a lesion or disease of the somatosensory sys-
of ‘‘normal’’ (‘‘physiological’’) pain and its mechanisms tems’’ [66]. This definition is the one we shall be following
of transmission, perception, and cortical control, but also here.
to the assessment of abnormal chronic pain — essentially A first-line observation is that neuro-imaging data,
of neuropathic origin. In the latter case, electrophysiology including functional imaging, have proved so far insufficient
uniquely contributes to diagnosis, and helps respond to one to predict, or even to establish, whether a given patient
most frequently asked clinical question, namely ‘‘is this a suffers or is susceptible to develop NP. Thus, while lateral
real neuropathic pain?’’ In a number of cases, neurophysi- medullary lesions (Wallenberg’s syndrome) can induce NP
ological studies may also provide information of prognostic in 25—50% of cases [47,58], whether or not a given patient
value for the development, or the evolution, of NP. Lastly, will develop pain after this type of lesion cannot be cur-
electrophysiological techniques may be also useful to guide rently established based on CT-, MRI or angiographic data.
therapeutic procedures, either by predicting their efficacy Because of the lack of a specific diagnostic tool for NP, a
or inefficacy, or by minimising the chances of pain devel- grading system of ‘‘definite’’, ‘‘probable’’, and ‘‘possible’’
opment through real-time neurophysiological monitoring, NP was recently proposed [66], whereby the diagnostic of
for instance during potentially nerve-injuring surgery. The ‘‘definite neuropathic pain’’ can be considered only in the
goal of neurophysiological pain assessment will be in general presence of:
terms to support, clarify, and quantify the patient’s subjec-
tive report, and in some cases to contradict such report if it • a neuroanatomically plausible distribution of pain;
appears incompatible with the objective neurophysiological • a history suggestive of somatosensory lesion/disease;
data. • at least one objective confirmatory test of the existence
Pain by excess of nociception (‘‘nociceptive pain’’) is of relevant somatosensory lesion or disease.
transmitted by an intact nervous system following stimula-
tion of nociceptors, while the so-called ‘‘neuropathic pain’’ Neurophysiological testing can provide such confirma-
(NP), is transmitted by an altered or diseased somatosensory tory test. The main available neurophysiological techniques
system. The classical definition of NP was for long a pain are, first, those of the standard electrophysiological exam-
caused by ‘‘lesion or dysfunction of the nervous system’’. ination: electroneurography with nerve conduction velocity
Objective pain diagnostics: Clinical neurophysiology 189
studies (NCV) and standard somatosensory evoked potentials therapeutic, as illustrated by the distinction between plexus
to electrical stimulation of large fibres (SEPs). Both tech- and root lesion in Fig. 1.
niques essentially explore non-nociceptive pathways: large Besides the high probability of mixed lesions in periphe-
myelinated (A) fibres in the periphery, and the dorsal col- ral trauma, some specific features of NP, in particular parox-
umn — medial lemniscus system at central level. Techniques ysmal pain, might be specifically linked to a functional
to explore specifically the nociceptive system (fibres A␦ impairment of large A fibres, probably demyelination, and
and C in the periphery, the spinothalamic system at central therefore can be associated with abnormalities in standard
level) are nociceptive flexion reflexes (‘‘RIII reflexes’’) and NCV and SEPs [68,69]. Association between paroxysmal pain
evoked potentials to radiant or contact-heat (laser-evoked and large fibre dysfunction is also supported by pain in
potentials, LEPs, or contact-heat evoked potentials, CHEPs). trigeminal neuralgia, linked to segmental demyelination at
Other techniques such as microneurography, which allows the root-brainstem entry zone, and by the ‘‘Lhermitte’s
investigating either large or thin fibres by introducing a nee- sign’’, closely associated with demyelination of spinal dor-
dle within the nerve itself [61], are also important, but not sal columns. The assessment of dorsal column function by
currently of clinical use and will not be dealt within this arti- SEPs has significant predictive value as to the efficacy or fai-
cle. Neither will be treated neurophysiological techniques lure of spinal cord stimulation (SCS) [63]. Indeed, abnormal
such as transcranial magnetic (TMS) or direct current (tDCS) or absent SEPs at central level (abnormal central conduc-
cortical stimulation, which may have therapeutic value in tion time) indicate degeneration of dorsal column pathways,
NP, but are not used as diagnostic tools. the activity of which appears essential for SCS efficacy [32].
Preoperative assessment of dorsal column function would
be for some authors more effective and have a lower rate
Standard investigations: electroneurography of false positive results than the classical percutaneous
and somatosensory evoked potentials test that is usually performed before definitive implanta-
tion [63]. Last but not least, in patients with central lesions
Considerable evidence has accumulated showing that NP the respective contribution of spinothalamic and lemniscal
is most especially associated with lesions of temperature system abnormalities might play a role in the development
and pain pathways, either peripheral [52,70], spinal [17] and characteristics of central pain [53,74].
or supraspinal [5,6,30], while the involvement of the large Given the above considerations, and the fact that stan-
fibres-dorsal-column-lemniscal system is not crucial for most dard NCV and SEP studies are readily obtainable in most
cases of NP. Inasmuch as standard nerve conduction studies neurophysiological departments, it follows that a standard
or SEPs do not explore the pain and temperature systems, neurophysiological assessment with NVC/SEPs should remain
but rather the non-nociceptive pathways, we might consider a first-line approach in case of suspected NP, before (or in
that their importance for the study of NP patients is very parallel to) more selective examinations of the pain and
limited. However, the study of NCV/SEPs remains of con- temperature pathways.
siderable importance in the evaluation of pain patients, for
several reasons. First, large and thin peripheral fibres are
anatomically mixed in nerves, plexuses and spinal roots, Specific assessment of nociceptive pathways
without spatial segregation up to the dorsal root entry
zone (DREZ, [62]), and therefore peripheral lesions, in par- Nociceptive reflexes
ticular traumatic or metabolic, tend to affect large and
thin fibres indistinctly. Mixed impairment of large and thin Spinal nociceptive reflexes are polysynaptic flexor responses
fibres in peripheral nerves can be used to detect impend- specifically related to nociception. These are obtained by
ing lesions during intraoperative monitoring. For instance, in peripheral stimulation of a sensory nerve, in general the
patients undergoing thoracotomy, non-specific nerve lesions sural nerve at the ankle or the cutaneous superficial radial
assessed by SEPs were correlated with the postoperative branch at the wrist, and recording the withdrawal motor
development of NP, which could be prevented by using response over a flexor muscle of same metameric level. The
muscle-sparing thoracotomy [4]. Second, NCV and SEP stud- label ‘‘RIII reflex’’ comes from the fact that, under certain
ies are readily available in most hospitals, provide robust conditions, the afferent branch of this reflex is formed by
and topographically precise data, and their abnormality to ‘‘type III’’ fibres of Lloyd classification (corresponding to
stimulation of a painful territory is objective evidence of A␦ fibres in Gasser and Erlanger’s nomenclature) [20,36]. In
somatosensory involvement — thus giving support to the practice, nociceptive reflexes are most commonly obtained
diagnosis of NP. In this domain, one important localising sign on the lower limbs, applying trains of electrical shocks to
is given by the proximal or distal location of the somatosen- the sural nerve and recording surface EMG over the ipsila-
sory lesion relative to the dorsal root ganglia. A proximal teral biceps femoris muscle [59]. Nociceptive flexor reflexes
lesion (between the ganglion and the spinal cord) allows can also be recorded in upper limbs, stimulating the median,
axonal flux to remain normal in the peripheral portion of ulnar or radial nerve, and recording the EMG over forearm
the corresponding nerves, distal to the ganglion, and there- muscles such as the flexor carpi ulnaris or radialis. How-
fore action potentials and peripheral nerve conduction are ever, the abundance of non-nociceptive reflexes in the upper
preserved when such nerves are stimulated, despite pro- limbs (reflexes linked to motor control) makes it difficult
found hypaesthesia. Conversely, lesions that are distal to the individualisation of purely nociceptive components, con-
dorsal root ganglia (neuropathies or plexopathies) reduce trary to lower limbs. The clinical use of nociceptive flexor
distal axonal flux and considerably alter NCV studies. This reflexes relies on the concordance between the threshold
may have important consequences, both diagnostic and of the reflex and that of a subjective painful sensation,
190 L. Garcia-Larrea
Figure 1 Somatosensory evoked potentials (SEPs) to non-painful stimulation of median nerves in a patient presenting with pain
and dysesthesiae in the right arm, involving the radial aspect of the forearm with non-systematised root distribution, following
a car accident. Plexus responses distal to ganglion are normal bilaterally (upper traces) while responses from dorsal horn and
caudal brainstem (2nd and 3rd rows) are attenuated and delayed to right median-nerve stimulation. Cortical responses (bottom
line), albeit bilaterally present, are also delayed after right median-nerve stimulation. These results show axonal loss and abnormal
transmission time between spinal ganglion and the spinal cord with a C6-C7 level on the right side. Although standard SEPs only
reflect conduction in large non-nociceptive fibres, these results demonstrate impairment of dorsal roots and/or the dorsal root
entry zone over a territory consistent with pain distribution, and therefore substantiate the diagnosis of neuropathic pain.
first demonstrated by Hugon [36] and Willer et al. [75]. during these procedures has been shown to be associ-
Thus, when a given stimulation is able to trigger a segmen- ated with a good clinical efficacy at short, and perhaps
tal nociceptive reflex in normal individuals, it is commonly middle-term [23,26]. The possible predictive value of this
also able to activate the ascending nociceptive pathways phenomenon in the long-term has not, however, yet been
and generate a subjective sensation of pain, and this allows determined.
characterising states of hyper- or hypo-algesia as well as
studying possible dissociations between reflex responses and
Nociceptive evoked potentials
subjective pain [59]. However, it should be underlined that
nociceptive reflexes are polysynaptic responses, which are
highly sensitive to descending, inhibitory or facilitatory, Given the significant link between the involvement of
controls; attention and distraction can drastically modify pain/temperature pathways and the development of NP, it is
response characteristics, so that very strict conditions of logical that much energy has been devoted in past years to
sustained attention are mandatory when these responses are develop stimulation procedures aiming at obtaining a selec-
used for clinical purposes. tive activation of thin fibres (A␦ and C) in the periphery and
Despite its theoretical advantages, the RIII reflex has spino-thalamo-cortical pain systems at central level. The
been so far little used as a diagnostic tool [59]. This next sections summarises this.
response can be useful in the diagnosis of small-fibre neu-
ropathy; indeed, the coexistence of normal standard NCV Noxious electrical stimuli
and preserved sensory thresholds to low-intensity stimuli,
but increased pain thresholds and attenuated or abolished Transcutaneous electrical stimulation at pain threshold is
nociceptive reflexes is a signature of this condition. Also, efficient to get nociceptive spinal reflexes, as the effer-
the recording of both RIII reflexes and concomitant pain ent reflex branch is specifically triggered when nociceptive
sensations in response to low-intensity stimuli is an objec- afferents are activated (see above). Conversely, such stimuli
tive sign of allodynia, while dissociation between preserved are clearly inadequate to obtain cortical responses reflect-
reflexes but abolished pain sensation suggest a spinothala- ing specifically the pain/temperature ascending volleys.
mic lesion rostral to the spinal level of reflex integration Indeed, whenever an electrical stimulus reaches painful lev-
[24]. The RIII reflex has also been used as a criterion of els, it stimulates concomitantly large and small periphe-
‘‘quality control’’ testing the efficacy of spinal or motor ral afferents, but the resulting cortical potentials are
cortex stimulation, whereby attenuation of the RIII reflex largely dominated by the non-nociceptive signals [16]. This
Objective pain diagnostics: Clinical neurophysiology 191
Figure 2 Laser-evoked potentials (LEPs) in a patient with neuropathic pain after brachial plexus avulsion. Abnormal LEPs after
stimulation of the painful dermatome (C6). Stimulation of the contiguous, affected but non-painful dermatome (C5), as well as
that of the contralateral side, yielded normal LEPs. The figure illustrates both the ability of LEPs to detect A␦-fibre impairment in a
single radicular territory, and the association of such impairment to the development of neuropathic pain (modified from ref. [14]).
phenomenon has been labelled the ‘‘first come, first [8]. However, current spread of such stimulations most
served’’ principle. It stipulates that when several groups of probably entails the simultaneous activation of A␦ fibres,
fibres of different conduction velocities reach the cortex which makes the stimulation non-selective. This is indeed
sequentially, cortical responses tend to reflect preferen- suggested by the very short-latency of scalp potentials
tially those arriving first [29,67]. This principle entails recorded with this method, which may be as low as 30—40 ms
that, in order to get evoked potentials that are spe- [35]. Inui and co-workers [38] introduced a new method
cific to nociceptive input, such input should be devoid of of intra-epidermal electrical stimulation using the tip of
any contamination by large non-nociceptive fibres. Elec- a stainless steel needle inserted 0.2 mm in the epidermis.
trical stimulation of special organs exclusively innervated This intra-epidermal technique did not produce any scalp
by small fibres, such as the tooth pulp, yields scalp response earlier than 100 ms, and estimated conduction
responses that, at least theoretically, are specific to noci- velocity of the activated fibres was about 15 m/s, thus con-
ception, and whose amplitude correlates positively with sistent with selective A␦ stimulation. However, recent data
the pain sensation [21,33]. The main drawbacks of these suggested that such selectiveness may disappear for sti-
methods are technical difficulty, confinement to just one mulus intensities approaching nociceptive ranges [51], thus
territory, and the fact that the tooth pulp may also limiting possible clinical application of the technique. The
contain some A fibres [12], perhaps explaining why ‘‘pre- same criticism holds true for the ‘‘planar concentric elec-
pain’’ sensations other than pinprick may be obtained trode’’, whish has also been recently described [42,43]:
with such stimulation [2]. Experimental blocks of large despite its clever design, this electrode also significantly co-
peripheral fibres by direct current can provide a selec- activates large A fibres [18], and failed to detect proven
tive nociceptive input [48,76]. Although this method is lesions of pain pathways [56].
certainly useful, it is not appropriate for clinical appli-
cation, as the large fibre blocks are only effective on
exposed nerves. Compression blocks are time-consuming Responses to contact-heat (CHEPs)
and restricted to few territories, mostly the superficial
radial nerve at the wrist. Intraneural microstimulation [64] Standard thermodes such as those used in functional ima-
is a powerful technique to stimulate selectively small- ging studies do not allow temperature changes rapid enough
diameter afferents. However, its technical difficulties, to evoke EPs (review in [55]). New generations of fast-
together with the duration of the test and its unpleasant acting heat foil thermodes permit temperature rise times
character, make this procedure hardly applicable in pain up to 70 ◦ C per second and can evoke cortical responses
patients, and mainly restricted to experimental settings [11,40]. Rapid thermode heating generally evokes a pinprick
[61]. sensation, and CHEPs have therefore been attributed to A␦
Intracutaneous stimulations were described as early as fibre stimulation, although CHEP latencies are delayed by
1984 to activate preferentially thin epidermal afferents 200—300 ms relative to the A␦ EPs obtained using laser pulses
192 L. Garcia-Larrea
Figure 3 This 65-year woman underwent a minor head trauma (temporal concussion, no loss of consciousness) and a CT-scan
performed in the emergency room was considered normal at supratentorial level. Four weeks later, she developed burning pain in
the right upper limb, which was not considered neuropathic despite a new CT-scan that showed an asymmetry at midbrain level (June
2006). Standard SEPs were normal. While malingering was being seriously considered, laser-evoked potentials (LEPs) demonstrated
significant abnormalities of conduction in pain-temperature pathways after stimulation of the painful right hand (bottom left) and
remained normal to stimulation of the non-painful side (top left). A gradient-echo (T2*) MRI showed a hyposignal image in the
posterior left midbrain, suggesting a haemorrhagic sequel. This small midbrain lesion impinged with spinothalamic conduction and
was responsible of the neuropathic pain.
[40]. Such differences suggest that the peripheral afferent Responses to radiant heat stimulation:
CHEP volley is triggered relatively late during the heating laser-evoked potentials (LEPs)
ramp; accordingly, single thalamic units appear to respond
up to 300 ms after the maximal thermode temperature has Most of the difficulties inherent to electrical or contact-
been reached [57]. The precise moment of afferent-fibre heat stimuli, described above, can be circumvented using
activation during the thermode heating ramp is hard to radiant heat stimulation with laser pulses. Laser beams pro-
establish, as it may change from one individual to another, vide a rapid, synchronous and selective activation of A␦
and from one trial to the following; this may explain the and C- thermosensitive nociceptors in the hairy skin, with-
variability reported in the CHEP latencies from different out any concomitant activation of mechano-receptors and
laboratories (see [10] vs. [11] vs. [40]). With further refine- A fibres. Although other sources of thermal stimulation
ments, the CHEP technique may become a useful tool to such as light bulbs or Xenon lamps also permit selective
study pain-related responses, as the technique appears safe activation of thermo-receptors, they cannot ensure their
and is applicable over most body territories. Substantial synchronous activation, and therefore are hardly useful for
work is however needed to standardise techniques, estab- recording neurophysiological responses. Furthermore, con-
lish the nature of the stimulated fibres, explain the sources ventional heat sources mainly emit in the visible spectral
of latency variability, and determine the possible influence range, where skin energy absorption is poor and widely
of mechanoreceptor activation by the contact probe [31]. affected by changes in skin pigmentation, thus making
Last but not least, methods should be devised to keep the control of heat transfer virtually impossible (review
constant the energy transfer, which depends on the pres- in [55]). Most if not all of these problems are avoided by
sure of a flat thermode against cutaneous surfaces that are using monochromatic high-intensity light sources such as
rarely flat. those provided by laser stimulators, of which gas (CO2 )
Objective pain diagnostics: Clinical neurophysiology 193
Figure 4 Post-thoracotomy neuropathic pain with selective abnormality of laser-evoked responses to stimulation of C-fibres. In
the upper part of the figure, standard laser stimulation using high energy laser pulses with small spot size (Nd:YAP; 80 mJ/mm2,
4 mm spot) yields normal and symmetrical responses from the painful and non-affected territories. In the lower panel, selective
stimulation of C-warmth fibres using lower energy densities and a larger spot size (30 mJ/mm2, 20 mm spot diameter) yielded
significantly attenuated C-LEPs (left) and sympathetic skin responses (right) to stimulation of the painful region exclusively. Please
note that the time bases are different for A␦ and C-fibre responses.
and solid-state (Th:YAG/Nd:YAP) laser stimulators are the and weak contribution of the first somatosensory areas. The
most commonly used. Solid-state lasers beams are easily ‘‘vertex complex’’, or ‘‘N2/P2’’, which has been the most
transported via optic fibres, rendering manipulation of the frequently used in the clinics, reflects a combined activa-
stimulation much easier than with CO2 lasers, and allowing tion of the middle cingulate and bilateral insulae, with some
stimulating virtually any body territory. contribution of parietal and prefrontal regions [22,28,72].
Energy density of infrared laser pulses increases local It is important to note here that the positive component of
temperature with a ramp up to 10 000 ◦ C/sec, which allows the vertex complex, or ‘‘P2’’, is a mixed wave loaded with
rapid and synchronised activation of free nerve endings of A␦ late cognitive components, the amplitude of which may vary
and C-fibres, and thus recording of LEPs. Depending on the widely with attention, vigilance and other cognitive factors
ramp and the temperature achieved, either both A␦ and C- [3,25,45,46]. Thus, for clinical purposes it is advisable to
fibres, or the latter alone can be activated. Inconveniences rely, whenever possible, on early responses N1 and N2, which
linked to this type of stimulation are the need of strict are less subject to cognitive modulation [25,46].
laboratory conditions, the possibility of skin lesions due to Although laser stimulation activates simultaneously A␦
manipulation errors (extremely rare with solid-state lasers). and C-fibres, the cortical responses that are obtained with
Besides, a common drawback to all nociceptive EPs, as standard techniques reflect almost exclusively the A␦ com-
compared with standard (A) SEPs, is the fact that only corti- ponent of the afferent volley. Selective excitation of the
cal responses can be recorded satisfactorily. Indeed, the low C-fibre component is however possible by a diversity of
number of stimuli delivered, together with the lack of spa- procedures. The earliest methods described were based on
tial summation and the high time-dispersion of A␦ afferent experimental pressure-block of group A fibres [9] or spec-
volleys in thin fibres, render the recording of peripheral or tral analysis of the expected time windows [1]. Since then,
spinal potentials almost inaccessible. Both source modelling a number of easier techniques have been proposed, such
and intracranial recordings have permitted to determine as the stimulation of tiny skin areas [7,54], selection of
the main cortical origin of LEPs recorded within the first single trials devoid of A␦-LEPs [65] or stimulation at low
400 milliseconds following nociceptive laser stimulation. intensity [13,49,67]. Each of these procedures is based on
The earliest responses (scalp N1/P1 or N160/P160) reflect different physiological principles. The use of tiny stimula-
activation of the operculo-insulair cortex (SII and poste- tion surfaces (0.15—0.30 mm2 ) for the selective activation
rior insula) in the suprasylvian region, with a simultaneous of C-fibres takes advantage of the higher density of C — with
194 L. Garcia-Larrea
respect to A␦ receptors (review in [55]), while the use of produce very discrete alterations of LEPs, thus underscoring
low stimulus intensities (below the A␦ activation threshold) the fact that spinothalamic terminals within the thalamus
stands on the fact that the heat threshold is lower in C than reach multiple sites including lateral and medial targets, as
in A␦ nociceptors. Purely warm sensations related to C-fibre well as the posterior suprageniculate group (including the
activation can be evoked by laser pulses that irradiate a thalamic region called VMpo). Dissociations between lemnis-
large skin area (50—200 mm2 ), and this appears the easi- cal and spinothalamic responses (SEPs and LEPs) have been
est technique so far to obtain C-evoked cortical potentials described in thalamic lesions [53].
[13,37,67]. ‘‘Ultra-late’’ LEPs related to C-fibre activation
culminate later than standard A␦ LEPs, the positive vertex
component peaking at 800—1200 ms following upper limb Conclusion no 3. Partial preservation and
stimulation. desynchronisation of laser-evoked potentials in
neuropathic pain increase the probability of
Conclusions allodynic phenomena
SEPs and LEPs are techniques easy to implement in any Although this conclusion would benefit from further valida-
laboratory of clinical neurophysiology. They reflect accu- tion, it is physiologically plausible and has been supported
rately and objectively the somatosensory transmission, by a number of independent studies so far. A study on
respectively within lemniscal (SEPs) and spinothalamic 54 patients reported that complete obliteration of LEPs to
pain/temperature pathways (LEPs). Their use is now beco- stimulation of the painful territory was associated with a
ming a ‘‘standard’’ to document the possible involvement very small proportion of allodynic or hyperalgesic pheno-
of somatosensory systems in patients with suspected NP mena, while the prevalence of these phenomena increased
(hence fulfilling the requirements for NP diagnosis, [66]). In significantly when LEP alteration was only partial, and in
particular, during the last 10 years LEPs have proved their particular when LEPs appeared desynchronised or included
ability to detect even minute lesions within nociceptive late abnormal components [27]. These data have been con-
pathways, including monoradicular lesions (Fig. 2) or very firmed by independent studies in patients with peripheral
small lesions at brainstem, thalamic or cortical level (Fig. 3). or spinal NP, in whom both LEPs and thermal thresholds
Results gathered during these years have led to a number of were more severely affected in patients with ongoing pain
conclusions regarding the utilisation of evoked potentials in than in those with provoked pain (allodynia/hyperalgesia)
pain medicine, summarised as follows. [19,34,70].
[19] Ducreux D, Attal N, Parker F, Bouhassira D. Mechanisms of [39] Isnard J, Magnin M, Jung J, Mauguière F, Garcia-Larrea L.
central neuropathic pain: a combined psychophysical and fMRI Does the Insula tell the brain that we are in pain? Pain
study in syringomyelia. Brain 2006;129:963—76. 2011;152:946—51.
[20] Ertekin C, Ertekin N, Karcioglu M. Conduction velocity along [40] Itskovich VV, Fei DY, Harkins SW. Psychophysiological and
human nociceptive reflex afferent nerve fibres. J Neurol Neu- psychophysical responses to experimental pain induced by
rosurg Psychiatry 1975;38:959—65. two types of cutaneous thermal stimuli. Int J Neurosci
[21] Fernandes de Lima VM, Chatrian GE, Lettich E, Canfield RC, 2000;105:63—75.
Miller RC, Soso MJ. Electrical stimulation of tooth pulp in [41] Julien N, Goffaux P, Arsenault P, Marchand S. Widespread pain
humans. I. Relationships among physical stimulus intensities, in fibromyalgia is related to a deficit of endogenous pain inhi-
psychological magnitude estimates and cerebral evoked poten- bition. Pain 2005;114(1—2):295—302.
tials. Pain 1982;14:207—32. [42] Katsarava Z, Ayzenberg I, Sack F, Limmroth V, Diener HC,
[22] Frot M, Magnin M, Mauguière F, Garcia-Larrea L. Human SII and Kaube H. A novel method of eliciting pain-related poten-
posterior insula differently encode thermal laser stimuli. Cereb tials by transcutaneous electrical stimulation. Headache
Cortex 2007;17:610—20. 2006;46(10):1511—7.
[23] Garcia-Larrea L, Sindou M, Mauguiere F. Nociceptive flex- [43] Kaube H, Katsarava Z, Kaufer T, Diener HC, Ellrich JA. New
ion reflexes during analgesic neurostimulation in man. Pain method to increase nociception specificity of the human blink
1989;39:145—56. reflex. Clin Neurophysiol 2000;111:413—6.
[24] Garcia-Larrea L, Charles N, Sindou M, Mauguiere F. Flexion [44] Kirby R, Chouchou F, Perchet C, Magnin M, Garcia-Larrea
reflexes following anterolateral cordotomy in man: dissocia- L. The relationship existing between cortical responses,
tion between pain sensation and nociceptive reflex RIII. Pain sympathetic activities and subjective ratings to nociceptive
1993;55:139—49. stimuli delivered by Nd: YAP laser. Neurophysiol Clin 2009;39:
[25] Garcia-Larrea L, Peyron R, Laurent B, Mauguiere F. Associa- 67—8.
tion and dissociation between laser-evoked potentials and pain [45] Legrain V, Guerit J, Bruyer R, Plaghki L. Attentional modu-
perception. Neuroreport 1997;8:3785—9. lation of the nociceptive processing into the human brain:
[26] Garcia-Larrea L, Peyron R, Mertens P, Gregoire MC, Lavenne F, selective spatial attention, probability of stimulus occurrence,
Le Bars D, et al. Electrical stimulation of motor cortex for pain and target detection effects on laser-evoked potentials. Pain
control: a combined PET-scan and electrophysiological study. 2002;99:21—39.
Pain 1999;83:259—73. [46] Lorenz J, Garcia-Larrea L. Contribution of attentional and cog-
[27] Garcia-Larrea L, Convers P, Magnin M, Andre-Obadia N, Peyron nitive factors to laser-evoked brain potentials. Neurophysiol
R, Laurent B, et al. Laser-evoked potential abnormalities in Clin 2003;33:293—301.
central pain patients: the influence of spontaneous and pro- [47] Macgowan DJL, Janal MN, Clark MC, Wharton RN, Lazar RM,
voked pain. Brain 2002;125:2766—81. Sacco RL, et al. Central post-stroke pain and Wallenberg’s
[28] Garcia-Larrea, Frot M, Valeriani M. Brain generators of laser- lateral medullary infarction: frequency, character, and deter-
evoked potentials. From dipoles to functional significance. minants in 63 patients. Neurology 1997;49:120—5.
Neurophysiol Clin 2003;33:279—92. [48] Manfredi M. Differential block of conduction of larger
[29] Garcia-Larrea L. Somatosensory volleys and laser-evoked fibres in peripheral nerve by direct current. Arch Ital Biol
potentials: first come, first served? (Editorial) Pain 1970;108:52—71.
2004;112:5—7. [49] Magerl W, Ali Z, Ellrich J, Meyer RA, Treede RD. C- and A␦-
[30] Garcia-Larrea L, Perchet C, Creac’h C, Convers P, Peyron R, fibre components of heat-evoked cerebral potentials in healthy
Laurent B, et al. Opeculo-insular pain (parasylvian pain): a human subjects. Pain 1999;82:127—37.
distinct central pain syndrome. Brain 2010;133:2528—39. [50] Mouraux A, Guerit JM, Plaghki L. Non-phase locked electroen-
[31] Greffrath W, Baumgärtner U, Treede RD. Peripheral and central cephalogram (EEG) responses to CO2 laser skin stimulations
components of habituation of heat pain perception and evoked may reflect central interactions between A partial partial
potentials in humans. Pain 2007;132:301—11. differential- and C-fibre afferent volleys. Clin Neurophysiol
[32] Guan Y, Wacnik PW, Yang F, Carteret AF, Chung CY, Meyer 2003;114:710—22.
RA, et al. Spinal cord stimulation-induced analgesia: electrical [51] Mouraux A, Iannetti GD, Plaghki L. Low intensity
stimulation of dorsal column and dorsal roots attenuates dorsal intra-epidermal electrical stimulation can activate
horn neuronal excitability in neuropathic rats. Anesthesiology A-delta-nociceptors selectively. Pain 2010;150:199—207.
2010;113:1392—405. [52] Mendell JR, Sahenk Z, Clinical practice. Painful sensory neu-
[33] Harkins SW, Chapman CR. Cerebral evoked potentials to nox- ropathy. N Engl J Med 2003;348:1243—55.
ious dental stimulation: relationship to subjective pain report. [53] Montes C, Magnin M, Maarrawi J, Frot M, Convers P, Mauguiere F,
Psychophysiol 1978;15:248—52. et al. Thalamic thermo-algesic transmission: ventral posterior
[34] Hatem SM, Attal N, Ducreux D, Gautron M, Parker F, (VP) complex versus VMpo in the light of a thalamic infarct with
Plaghki L, et al. Clinical, functional and structural deter- central pain. Pain 2005;113:223—32.
minants of central pain in syringomyelia. Brain 2010;133: [54] Opsommer E, Guérit JM, Plaghki L. Exogenous and endogenous
3409—22. components of ultra late (C-fibre) evoked potentials following
[35] Howland EW, Wakai RT, Mjaanes BA, Balog JP, Cleeland CS. CO2 laser stimuli to tiny skin surface areas in healthy subjects.
Whole head mapping of magnetic fields following painful elec- Neurophysiol Clin 2003;33:78—85.
tric finger shock. Cognitive Brain Res 1995;2:165—72. [55] Plaghki L, Mouraux A. How do we selectively activate
[36] Hugon M. Reflex and perceptive effects of electric stimulation skin nociceptors with a high power infrared laser? Physiol-
of the sural nerve. Actual Neurophysiol (Paris) 1972;9:23—46. ogy and biophysics of laser stimulation. Neurophysiol Clin
[37] Iannetti GD, Truini A, Romaniello A, Galeotti F, Rizzo C, Man- 2003;33:269—77.
fredi M, et al. Evidence of a specific spinal pathway for the [56] Perchet C, Frot M, Chamarty A, Flores C, Mazza S, Magnin
sense of warmth in humans. J Neurophysiol 2003;89:562—70. M, Garcia-Larrea L. Do we activate specifically somatosensory
[38] Inui K, Tran TD, Qiu Y, Wang X, Hoshiyama M, Kakigi R. thin fibres with the concentric planar electrode? A scalp and
A comparative magnetoencephalographic study of cortical intracranial EEG study. Pain 2012, in press.
activations evoked by noxious and innocuous somatosensory [57] Pralong E, Pollo C, Bloch J, Villemure JG, Daniel RT, Tetreault
stimulations. Neuroscience 2003;120:235—48. MH, et al. Recording of ventral posterior lateral thalamus
Objective pain diagnostics: Clinical neurophysiology 197
neuron response to contact-heat-evoked potential in patient [67] Truini A, Galeotti F, Cruccu G, Garcia-Larrea L. Inhibition of
with neurogenic pain. Neurosci Lett 2004;367:332—5. cortical responses to A-delta inputs by a preceding C-related
[58] Ruel JH, Vighetto A, Confavreux C, Trillet M, Aimard G. Les response: testing the ‘‘first come, first served’’ hypothesis of
séquelles doulouresues du syndrome de Wallenberg. Presse Med cortical laser-evoked potentials. Pain 2007;131:341—7.
1992;21:926—8. [68] Truini A, Galeotti F, Haanpaa M, Zucchi R, Albanesi A, Biasiotta
[59] Sandrini G, Serrao M, Rossi P, Romaniello A, Cruccu G, Willer A, et al. Pathophysiology of pain in postherpetic neuralgia: a
JC. The lower limb flexion reflex in humans. Prog Neurobiol clinical and neurophysiological study. Pain 2008;140:405—10.
2005;77:353—95. [69] Truini A, Padua L, Biasiotta A, Caliandro P, Pazzaglia C, Galeotti
[60] Schestatsky P, Valls-Solé J, Costa J, León L, Veciana M, Chaves F, et al. Differential involvement of A-delta and A-Beta fibres
ML. Skin autonomic reactivity to thermo-algesic stimuli. Clin in neuropathic pain-related to carpal tunnel syndrome. Pain
Auton Res 2007;17:349—55. 2009;145:105—9.
[61] Serra J, Campero M, Bostock H, Ochoa J. Two types of [70] Truini A, Biasiotta A, La Cesa S, Di Stefano G, Galeotti F,
C nociceptors in human skin and their behavior in areas Petrucci MT, et al. Mechanisms of pain in distal symmetric
of capsaicin-induced secondary hyperalgesia. J Neurophysiol polyneuropathy: a combined clinical and neurophysiological
2004;91(6):2770—81. study. Pain 2010;150:516—21.
[62] Sindou M, Quoex C, Baleydier C. Fibre organization at the [71] Valeriani M, de Tommaso M, Restuccia D, Le Pera D, Guido M,
posterior spinal cord-rootlet junction in man. J Comp Neurol Iannetti GD, et al. Reduced habituation to experimental pain in
1974;153:15—26. migraine patients: a CO(2) laser-evoked potential study. Pain
[63] Sindou MP, Mertens P, Bendavid U, Garcia-Larrea L, Mau- 2003;105:57—64.
guiere F. Predictive value of somatosensory-evoked potentials [72] Valeriani M, Truini A, Le Pera D, Insola A, Galeotti F, Petrachi
for long lasting pain relief after spinal cord stimulation: C, et al. Laser-evoked potential recording from intracerebral
practical use for patient selection. Neurosurgery 2003;52: deep electrodes. Clin Neurophysiol 2009;120:790—5.
1374—83. [73] Veciana M, Valls-Solé J, Schestatsky P, Montero J, Casado V.
[64] Torebjork HE, Schady W, Ochoa JL. A new method for demon- Abnormal sudomotor skin responses to temperature and pain
stration of central effects of analgesic agents in man. Neurol stimuli in syringomyelia. J Neurol 2007;254:638—45.
Neurosurg Psychiatry 1984;47:862—9. [74] Wessel K, Vieregge P, Kessler C, Kompf D. Thalamic
[65] Towell AD, Purves AM, Boyd SG. CO2 laser activation of noci- stroke: correlation of clinical symptoms, somatosensory-
ceptive and non-nociceptive thermal afferents from hairy and evoked potentials, and CT findings. Acta Neurol Scand
glabrous skin. Pain 1996;66:79—86. 1994;90:167—73.
[66] Treede RD, Jensen TS, Campbell JN, Cruccu G, Dostrovsky [75] Willer JC. Comparative study of perceived pain and nociceptive
JO, Griffin JW, et al. Neuropathic pain: redefinition and a flexion reflex in man. Pain 1977;3:69—80.
grading system for clinical and research purposes. Neurology [76] Zimmermann M. Dorsal root potentials after C-fibre stimula-
2008;70:1630—5. tion. Science 1968;160:896—8.