Detection and prediction of neuropathic pain:the role of neurophysiological techniques

Luis Garcia-Larrea

NeuroPain lab - Central Integration of Pain in HumansInserm U1028 – Cnrs UMR5092 – Univ. Claude Bernard Lyon 1

University Hospital Pain Centre (CETD)Hôpital Neurologique Lyon, France

Clinical Neurophysiology in :

• Diagnosis

• Prediction

• Detection of pain states

Diagnosis

Chronic pain

nociceptiveneuropathic

Diagnosis of definite neuropathic pain needs:

1. neuroanatomically plausible distribution of pain, 2. history suggestive of a somatosensory lesion / disease, and 3. at least one objective confirmatory test of the existence of “relevant

somatosensory lesion or disease”

Neuropathic pain (IASP SIG 2008) :« Pain resulting as a direct consequence of a lesion or a disease of somatosensory systems »

dysfunctional

Treede, Jensen et al, Neurology 2008

Treede, Jensen et al, Neurology 2008

We need markers

Pain No pain

Images are not enough

Pain? Sensory deficit?

Images are not enough

https://www.theguardian.com/media/2017/jun/15/

abc-journalist-adam-harvey-hit-in-neck-by-stray-bullet-while-reporting-in-philippines

It literally just looks like I’ve been hit with a cricket ball”.

“I’ve got to go, but I’m OK.”

a) Definite diagnosis of Neuropathic pain needsobjective demonstration of lesion or diseaseof somatosensory pathways

b) Neurophysiology provides such objectiveindicators

Peripheral side Central side

• Aβ-mediated• electrical stimulus (SEPs)• tactile stimulus (t-SEPs)• air puff (a-SEPs)

• Aδ-mediated• laser stimulus (LEPs)• contact heat/cold stimulus (CHEPs)• intraepidermal electric currents (iEPs)

• C-fibre mediated• laser stimulus (C-LEPs)

Dorsal columns –medial lemniscus

Spino-thalamicsystem (STS)

… and multiple stimulus modes to activate nociceptive and non-nociceptives pathways

Modified from Treede, Lorenz & Baumgartner, Neurophysiol Clin 2003, 6: 303-14

Neuropathic pain is most often the

result of lesions in thin peripheral

fibres, and/or spino-thalamo-

cortical pathways.

Truini, Garcia-Larrea, Cruccu; Nat Neurol Reviews 2013

Nociceptive (RIII) reflexes

Sympathetic skin

responses

0 10 sec

Voluntary motor reactions

Multiple techniques at our disposal

Laser, contact heat/cold, intraepidermal

How can we stimulate systems for pain & temperature in the clinics?

Electrical Thermal Mechanical ChemicalStandard intraepidermal thermode light bulb laser

Specificity --- ++? ++ +++ +++ --- +++

Noninvasiveness +++ +++ +++ +++ +++ +++ ---

Reproducibility +++ ++ ± ± ++ ± ±

Time locking +++ +++ -- --- ++ ± ---

Adapted from an original idea of Léon Plaghki

N2

P2

N1

N/P1-bip

P1

Cz-nose

T4-Fp1

T4-nose

Fp1-nose

0 200 400 600 800ms

early response (N1:150-180 ms)

«Vertex response» (N2-P2:200-350ms)

3. Sympathetic skinresponse (nociceptive stim)

VAS = 0-10/102. Subjective ratings

4. Motor reaction time (250-300ms)

Four pillars leading to interpretation

1. Evoked responses

Laser-evoked potentials detect abnormal transmisison in thermo-algesic pathways

(thin fibres, spinothalamic tracts)

…which pass unnoticed to conventional CN techniques (EMNG, standard SEPs)…

Woman, 60y

Syringomyelia decompressed 15 years before

10 year after operation burning feet

No lemniscal signs

Tendon reflexes present symmetrical

Strandard EMG / ENG normal

LEPs, stim L5 (foot)

CZ-CZ-

-0.10 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90

+25 uV

Average Controls

Patient (Right/Left)

+25 uV

CZ-CZ-

-0.10 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90

+25 uV

Average Controls

Patient (Right/Left)

+25 uV

Small fibre neuropathy

L3P257

Cruccu & Garcia-Larrea, Clin Neurophysiol Suppl., 2004

Peripheral / radicular lesions

Montes et al Pain 2005

Fz

Cz

Pz

C3 C4

240ms 306ms

T4-Fz

T3-FzLateralised negativity

Focal thalamic lesions

Insular – opercular lesions

Normal short-latency somatosensory evoked

potentials (SSEPs)

N20

P45

0 20 40 60 80 100ms

3µV

+

_

F7* F3* FZ* F4* F8*

FT7* FC3* FCZ* FC4* FT8*

T3* C3* CZ* C4* T4*

TP7* CP3* CPZ* CP4* TP8*

T5* P3* PZ* P4* T6*

O1* O2*A1 (TP9)* A2 (TP10)*

FCZ*

CZ*

CPZ*

normal side

affected

side

SSR

Abnormal nocieptive laser evoked potentials (LEPs)

0 400 800ms

10µV

+

_

Lesions of nerve branches

• often iatrogenic

• often non recognised

Thoracotomy

inguinal hernia repair

Knee arthroplasty

Marchettini 2003

L3P0293

LEP

RCS

LEP

RCS

LEP

RCS

Genito-femoral lesion or malingering ?

Stim laser Saph int D

DOB…L3P302

Motor reaction

Sympathetic skin response

Cortical evoked

potentials (LEPs)

Stim laser Saph int G

-1000 uV

0

500 uV

0 4 8 sec 0 4 8 sec

0

Knee surgery 9 years before. Persisting pain and hypo/dysesthesiae.

Still in conflict with surgeon

We should also explore large fibres using standard techniques

(NCV, SEP)

Exploring thin fibres / spinothalamic system in NP is crucial

BUT

« (…) electric shock-like sensations probably arise from high-frequency bursts generated in demyelinated non-

nociceptive Aβ fibres.

Paroxysmal pain following implantation of a spinal stimulator… for pain relief !

«Lhermitte’s sign » type of pain

Neurophysiological techniques detect abnormalities in somatosensory transmission

• In both small and large peripheral fibres• In both lemniscal and spinothalamic pathways • Even due to minute lesions

Message

Abnormal somatosensory responses to stimulation of a painful region are a neurophysiological signature of neuropathic pain

Medico-legal value

II. Predicting pain

Clinical neurophysiology allows characterising sensory abnormalities in patients who are already in pain.

Could these techniques predict the occurrence of pain in patients who have not yet developed painful symptoms ?

Conditions able to induce chronic neuropathic pain* Brain or spinal stroke

* Spinal injury

* Multiple sclerosis (MS)

* Syringomyelia / syringobulbia

* Tumours, abscesses

* Systemic disease (Lyme, Lupus, Syphilis…)

* Peripheral neuropathy (diabetes, ischaemia, trauma, MNM)

* Plexus avulsion

* radiculopathy

• …

…and some ideas about prevalence

•Central post-stroke pain:

•8% (Andersen et al 1995, n>200)

•18% if somatosensory signs (Hansson 2004)

• 25-30% if thalamus involved (Bogousslavsky et al)

•MS: 28% (Boivie 1999, n=372)

•Syringomyelia : 67% (Attal et al 1999, n=18)

•Wallenberg : 27-50% (MacGowan et al 97, n=63)

•Anterolateral cordotomy : <10% !! (Tasker 1990)

Thalamic stroke: ~25% of patients will develop pharmaco-resistant thalamic pain

How to detect patients at risk ?

Montes et al Pain 2005

Fz

Cz

Pz

C3 C4

240ms 306ms

T4-Fz

T3-FzLateralised negativity

A clinical certitude: central pain most often concomitant with loss of pain-temperature sensations (Boivie et al 1988)

Anatomical data: posterior thalamic nuclei are predominantly affected in thalamic pain

A 9

R

STh

SN

RN

CL

mc

Pf CeM

CM

LD

VLpd

VLpl

VM

pl

MDpc

VLpv

VPMpc

A 7.2

R

LP

RN

SN

CM

Pf

mc

CL

LD

STh

pl

MDpc

VLpv

VPLa

VPM

VPMpc

A 5.4

R

LGN

SN

RN

LP

Pf

CM

CL

PuA

LD

ZI

MDpc

VPM

VPLp

VPMpc

VPLa

CL

A 3.6

R

LGN

SN

RN

LP

PuA

VPM

VPI

CM

Pf

PG

Pf

VPLp

MDpc

VPMpc

A 2.7

R

LGN

PG

SN

RN

LP

PuA

CM

Pf

Li

CLPuM

Hb

Po

MD

VPLp

VPIVMpo

A 5.4

A 1.8

R

LP

VPI

PoPf

CMPuA

PuM

CL

Hb

RN

SN

LGN

Li

SGVPL

p

VMpo

VMpo unaffected

VPL-VPM-VPI-PuA

VMpo

VMpo

VPL

VPL

VLp

VPM

A 9

R

STh

SN

RN

CL

mc

Pf CeM

CM

LD

VLpd

VLpl

VM

pl

MDpc

VLpv

VPMpc

A 9

R

STh

SN

RN

CL

mc

Pf CeM

CM

LD

VLpd

VLpd

VLpl

VLpl

VM

pl

MDpc

MDpc

VLpv

VLpv

VPMpc

VPMpc

A 7.2

R

LP

RN

SN

CM

Pf

mc

CL

LD

STh

pl

MDpc

VLpv

VPLa

VPM

VPMpc

A 7.2

R

LP

RN

SN

CM

Pf

mc

CL

LD

STh

pl

MDpc

MDpc

VLpv

VLpv

VPLa

VPLa

VPM

VPMpc

VPMpc

A 5.4

R

LGN

SN

RN

LP

Pf

CM

CL

PuA

LD

ZI

MDpc

VPM

VPLp

VPMpc

VPLa

A 5.4

R

LGN

SN

RN

LP

Pf

CM

CL

PuA

LD

ZI

MDpc

MDpc

VPM

VPLp

VPLp

VPMpc

VPMpc

VPLa

VPLa

CL

A 3.6

R

LGN

SN

RN

LP

PuA

VPM

VPI

CM

Pf

PG

Pf

VPLp

MDpc

VPMpc

CL

A 3.6

R

LGN

SN

RN

LP

PuA

VPM

VPI

CM

Pf

PG

Pf

VPLp

VPLp

MDpc

MDpc

VPMpc

VPMpcposterior

anterior

A 2.7

R

LGN

PG

SN

RN

LP

PuA

CM

Pf

Li

CLPuM

Hb

Po

MD

VPLp

VPIVMpo

A 2.7

R

LGN

PG

SN

RN

LP

PuA

CM

Pf

Li

CLPuM

Hb

Po

MD

VPLp

VPLp

VPIVMpo

A 5.4A 5.4

A 1.8

R

LP

VPI

PoPf

CMPuA

PuM

CL

Hb

RN

SN

LGN

Li

SGVPL

p

VMpo

A 1.8

R

LP

VPI

PoPf

CMPuA

PuM

CL

Hb

RN

SN

LGN

Li

SGVPL

pVPL

p

VMpo

VMpo unaffected

VPL-VPM-VPI-PuA

VMpoVMpo

VMpoVMpo

VPLVPL

VPL

VLpVLp

VPMVPM

*receiving STT projections*affected by the infarct

Combining MRI and spinothalamic-evoked potentials

to classify patients with / without thalamic pain

following a thalamic lesion

Thalamic stroke with or without thalamic pain

(n=42 patients)

Vartiainen et al, Brain 2016

Phenotyping patients to predict / prevent neuropathic pain

Pain No Pain

LEP suppression + PuA involvement 80-90% thalamic pain development(see also Sprenger et al Brain 2012, and Kurtz et al JNNP 2012))

“Both abnormal LEP and involvement of the

anterior pulvinar nucleus were independently

associated with the development of thalamic pain.

The combination of both indices increased their

predictive value (PPV 87%, NPV 77.7%)”

Much more difficult:

Discriminating pain patientsamong those having undergone a selective spinothalamic lesion

Wallenberg syndrome: 25-50% of cases will develop central pain

ms-100.0 0.0 100.0 200.0 300.0 400.0 500.0 600.0 700.0 800.0 900.0

µV 0.0

-2.5

-5.0

-7.5

-10.0

2.5

5.0

7.5

10.0

__ normal side

__ affected side, non painful (n=9)

__ affected side, painful (n=11)

Different LEP abnormalities after a brainstem lesion

Szy

Valo

Ama

neuropathic pain

Normal

Affected

Time-frequency transform of LEPs

Controls

Wallenberg – non pain

Wallenberg - pain

Fré

quency

time

controls

Wall, no pain

Wall, pain

s

3.A

3.B

AMAX

AMAX

ΔT/ΔF

1/ [ΔT/ΔF]

Controls

Non-pain patients

Pain patients

Predicting pain in brainstem syndromes ?

III. Objective detection of a pain state

…using bodily reactions

Sympathetic responses to assess the arousal effectof nociceptive stimuli

Heart rate

Chouchou et al, EJP 2011

RR intervals (ms)

Pupillary Diameter (cm)

Systolic Pressure

Diastolic Pressure

Sympathetic skin response (µS, %)

Chouchou et al, submitted

Sympathetic skin responses reflect the arousal effectof nociceptive stimuli

Abnormal sudomotor skin responses to temperature and

pain stimuli in syringomyelia .

Veciana M, Valls-Solé J, Schestatsky P, Montero J, Casado V.

J Neurol. 2007 May;254(5):638-45

Skin autonomic reactivity to thermoalgesic stimuli

Schestatsky P, Valls-Solé J, Costa J, León L, Veciana M, Chaves ML..

Clin Auton Res. 2007;17:349-55.

Enhanced subjective pain is associated wth higher sympathetic skin responses

•Laser stimuli do not produce idential sensation in every single trial (changes in

energy delivery, receptor density, etc)

•Laser stimuli yielding higher SSR were associated with Higher BOLD responses in

the insula and S1/S2, and with higher amplitudes of laser-evoked cortical responses

Mobascher et al, NeuroImage 2009; 44: 1081-1092

Cotation contre RCSga

R2 = 0,5498

0

1000

2000

3000

4000

5000

6000

7000

8000

9000

0 1 2 3 4 5 6

Cotation

Am

plitu

de

de

RC

Sg

a

The P value is < 0.0001, considered extremely

significant.

Cotation contre N2P2 R2 = 0,3953

-20

0

20

40

60

80

100

0 1 2 3 4 5 6

Cotation

Am

plitu

de

de

N2

-P2

The P value is < 0.0001, considered

extremely significant.

sec-1.0 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0

0 900ms

Subject: Math…

VAS=5

VAS=2

VAS=5

VAS=2

In healthy subjects subjective intensity, cortical response and SSR covary

-1000 0 5000 9000

1.5J hey, it hurts!

1.25J stronger!

1J yes, it pricks

0.75J perhaps?

0.5J no resp

-3000

-2000

-1000

0

1000

2000

3000

Defining objective thresholds using sympathetic responses

Mme Cap… L3P0490

Objectivising hypoesthesia using sympathetic cutaneous responses

-1000 uV

0

500 uV

0 4 8 sec

-1000 uV

0

500 uV

0 4 8 sec

Stim Right T5; energy=1J Ø= 4mm

« very clear, pricking »

Stim Left T5; energy: 1J Ø= 4mm

« was there anything? »

Laser stim on the back,

SSR recorded over palm of the hand

Pain & hypoesthesia over left T5/T6, following back surgery + chiropractic manipulation

Objective assessment of thermal hyperalgesia

Patient L3P0492

FC1

0

FC2

0

C3

0

C4

CP2

0

P4

00

P3

CP1

Cz

0

0 400 800ms

Pz

0

Probable lesion right femorocutaneous n.Normal tactile (FVF) and thermal (laser) thresholdsMechanical allodyniaThermal hyperalgesia

Higly abnormal cortical responses

to laser stimulus of the painful side (red)

RT (EMG)

0

Abnormally delayed motor response times

Thermal hyperalgesia: sequential sympathetic responses

L3P0492

0 5000 9000 ms

-3000

-2000

-1000

0

1000

2000

3000

Average n=10 SSR

Probable surgical lesion right femorocutaneous n.Normal tactile (FVF) and thermal (laser) thresholdsMechanical allodyniaThermal hyperalgesia

Sequentially obtained SSR. Superimpsed traces to stim of right (red) and left FC territory

-1000 0 5000 9000

-3000

-2000

-1000

0

1000

2000

3000

Stim painful side:

thermal hyperalgesia with enhanced SSR

Stim normal side

Ham… Sa L3P047..

Traumatic lesion, right foot. Allodynic sequelae (mechanical & thermal). Clinical exam

impossible due to pain to any cutaneous contact. Doubt about veracity of allodynic symptoms

Laser evoked potentials impossible (intensity needed to record could not be reached)

Detecting allodynia using sympathetic responses

RCS Avg-chevdr-iee-RCS

-1000-800-600-400-200020040060080010001200140016001800200022002400260028003000320034003600380040004200440046004800500052005400560058006000620064006600680070007200740076007800800082008400

-800

-600

-400

-200

0

200

400

600

800

Intraepidermal stimuli, low intensity (0.2 mA)

Stim Right foot:«ouch, painful !!»

Stim Left foot:barely perceived

Anaesthesiol Intensive Ther. 2013 Jul-Sep;45(3):134-7.

Objective assessment of pain-related stress in mechanically ventilated newborns based on skin conductance fluctuations.

J Karpe, A Misiołek, A Daszkiewicz, H Misiołek.

Differentiating between heat pain intensities: the combined effect

of multiple autonomic parameters.

R Treister, M Kliger, G Zuckerman, I Goor Aryeh, E Eisenberg.

Pain 153 (2012) 1807-14.

Differentiating between heat pain intensities: the combined effect

of multiple autonomic parameters.

R Treister, M Kliger, G Zuckerman, I Goor Aryeh, E Eisenberg.

Differentiating between heat pain intensities: the combined effect

of multiple autonomic parameters.

R Treister, M Kliger, G Zuckerman, I Goor Aryeh, E Eisenberg.

Pain 153 (2012) 1807-14.

Sympathetic electrodermal activityto assess pain in the clinics

Sympathetic electrodermal activityto assess pain in veterinary medicine

“Lameness and pain were assessed using a visual analogue scale (VAS),

numerical rating scale (NRS), and electrodermal activity (EDA)”

“.. Electrodermal activity (EDA) was measured to assess objective pain”

Robocop …?

arterial pressure

EEG 128 chann

Pupillometrics

EKG

Painful

Stimuli

Subjective pain

assessment

Electrodermal activity

Detection of altered somatosensory transmission-> essential for doagnosis; medico-legal value

Prediction of neuropathic pain after neural lesion -> seems possible (thalamic stroke, brainstem lesion…)

but urgent need of prospective studies

Objectivation of pain combining behavioural, cortical and vegetative responses

-> Useful to detect objectively perceptive thresholds -> Probably useful to detect and follow allodynia (need more data!)

Why neurophysiology?

A conclusion needed ?

What is simple is false; what is complex is unusable

(Paul Valéry)

All this was done with :

Caroline Perchet

Philippe Convers

Michel Magnin

Roland Peyron

Maud Frot

François Mauguière

Patrick Mertens

Bernard Laurent

Christelle Creac’h

…