Formation Post-Universitaire

1266

Rev Neurol (Paris) 2007 ; 163 : 12, 1266-1270

G. LAURIA

ActualitésXI

es

Journées des Maladies du Système Nerveux Périphérique. ControverseRecent developments in the management of peripheral neuropathy using skin biopsy

G. LauriaAdresse :

Neuromuscular Diseases Unit, National Neurological Institute “Carlo Besta”, Milan, Italy.

Correspondance :

G. L

AURIA

, Neuromuscular Diseases Unit, National Neurological Institute Carlo Besta, Via Celoria, 11, 20133,Milan, Italy. E-mail: glauria@istituto-besta.it

SUMMARY

Skin biopsy has become a widely used tool to investigate small calibre nerve fibres in peripheral neuropathies.This technique is safe, minimally invasive, painless, easy to perform, and cheap. It provides diagnostic informa-tion in patients with small fibre neuropathy in whom routine neurophysiological tests are commonly normal.Moreover, it allows investigating the innervation of sweat glands, thus giving information on the autonomicnervous system. Biopsy of the hairy skin is used to investigate unmyelinated and small myelinated fibres, whe-reas biopsy of the glabrous skin can be taken to examine large myelinated fibres. The applications of skinbiopsy for diagnostic and research purposes cover the spectrum of peripheral nervous system diseases, frompainful axonal neuropathies to sensory neuronopathies and immune-mediated and inherited demyelinatedneuropathies. Finally, studies on axon regeneration in human and experimental models suggest that skin biop-sy has a potential usefulness to monitor the progression of neuropathy and the efficacy of neuroprotectivetreatments.

Keywords:

Skin biopsy • Peripheral neuropathy • Painful neuropathy • Autonomic neuropathy • Neuropathicpain

RÉSUMÉ

La biopsie cutanée : les progrès d’un outil diagnostique des neuropathies périphériques.

G. Lauria, Rev Neurol (Paris) 2007; 163: 12, 1266-1270

La biopsie cutanée est largement utilisée pour dépister les neuropathies périphériques à petites fibres. Cettetechnique peu invasive et indolore est facile à réaliser et peu coûteuse. Elle permet le diagnostic chez les patientsprésentant une neuropathie à petites fibres où les résultats obtenus par l’ENMG peuvent être normaux. Parailleurs, elle permet l’étude de l’innervation des glandes sudoripares et fournit ainsi des informations sur le sys-tème nerveux autonome. La biopsie cutanée en zone pileuse permet l’étude des fibres amyéliniques et des peti-tes fibres myélinisées, tandis que la biopsie en zone glabre renseigne sur les grandes fibres myélinisées. Lesapplications cliniques et celles de recherche sont nombreuses pour l’ensemble des neuropathies périphériques,allant des neuropathies axonales douloureuses aux neuronopathies sensitives et aux neuropathies démyélinisan-tes dysimmunitaires et héréditaires. Enfin, l’étude de la régénération axonale chez l’homme et chez l’animal sug-gère que la biopsie cutanée peut contribuer au suivi de l’évolution des neuropathies et de l’efficacité destraitements neuroprotecteurs.

Mots-clés :

Biopsie cutanée • Neuropathie périphérique • Neuropathie douloureuse • Neuropathie du sys-tème nerveux autonome • Douleur neuropathique

© 2007. Elsevier Masson SAS. Tous droits réservés

Actualités •

Skin biopsy

1267

G. LAURIA

Innervation of the human skin

The existence of a rich innervation in theepidermis and sub-papillary dermis of hu-man beings was definitely demonstratedusing antibodies against the protein geneproduct 9.5 (PGP 9.5), a neuronal form ofubiquitin carboxyl terminal hydrolasetransported by the slow component of axo-nal transport and widely expressed in theperipheral nervous system (Dalsgaard

etal,

1989). Biopsy of the hairy skin is mostcommonly performed to investigate un-myelinated nerve fibres. Hairy skin coversmost part of the body, including the limbs.Its epidermis is innervated by unmyelina-ted free-endings with exclusive somaticfunction, which are strongly immuno-reactive to PGP 9.5. This was demonstra-ted by their degeneration after experimen-tal axotomy or dorsal root ganglia (DRG)lesion, but not after dorsal rhizotomy orsympathectomy (Li

et al.,

1997). Intra-epidermal nerve fibres (IENF) arise fromsubpapillary dermal nerve bundles aftercrossing the dermal-epidermal junction.Most dermal fibres are unmyelinated andincluded in Remak bundles, but also somemyelinated fibres can be found. IENF runtoward the skin surface without enteringthe stratum corneum, with linear course,few branching, and slight varicosities li-kely due to the uneven distribution of thecytoplasmatic components. IENF have thepeculiar characteristic to be naked axons.In fact, subpapillary nerve fibres, fromwhich they arise, loose the Schwann cellensheathment while crossing the dermal-epidermal junction (Lauria

et al.,

2004).The density of IENF decreases from theproximal to the distal sites of the body, beinghigher at the paravertebral regions of the tru-nk than at the extremities and 60 p.cent hi-gher in the proximal thigh than in the distalregion of the leg (Lauria

et al.,

1999). Thereason of such distribution is still unknown.Some studies showed that aging affects thedensity of IENF, but this finding needs to beconfirmed (Lauria

et al.,

1999; Gøransson

etal.,

2004; Pan

et al.,

2001; Chien

et al.,

2001; Shun

et al.,

2004; McArthur

et al.,

1998). IENF can be labelled by antibodiesagainst specific components of the cytoske-leton, mainly microtubules (Lauria

et al.,

2004), and axonal membrane epitopes (Po-lydefkis

et al.,

2004), and widely express the

capsaicin receptor (Ständer

et al.,

2004; Lau-ria

et al.,

2006). Conversely, they are scar-cely immunoreactive to neuropeptides,which are mainly expressed by autonomicnerve fibres. IENF are involved in pain andthermal sensation transduction, but likelyhave further functions which have been onlypartly explored. Most of them are nocicep-tors and some have synaptic-like contactswith Langerhans’ cells, which are immune-competent cells, and keratinocytes, whichalso express thermal receptors, suggests theexistence of a complex nerve-cellnetwork.Biopsy of the glabrous skin can be obtai-ned from the fingers and the forearms andallow investigating myelinated fibres andsensory corpuscles, including Merkel cellsand Meissner corpuscles (Nolano

et al.,

2003). Pacini and Ruffini corpuscles resi-de deeply in the dermis and are commonlyexcluded from routine examination. It hasbeen recently demonstrated that the struc-ture of large myelinated nerves in the gla-brous skin is similar to that elsewhere inthe peripheral nervous system, and thatultrastructural and protein expression stu-dies can be easily performed in inheriteddemyelinating neuropathies (Li

et al.,

2005; Sabet

et al.,

2006).Punch biopsy allows investigating alsodermal nerve fibres, including those inner-vating the autonomic organs. Sympatheticcholinergic fibres of sweat glands, andnerves of arrector pilorum muscles, hairfollicles, and vessels can be immunostainedby antibodies against PGP 9.5, and neuro-peptides, such as calcitonin gene relatedpeptide, substance P, and vasointestinalpeptide. Several studies (Pan

et al.,

2001;Karanth

et al.,

1989; Kennedy

et al.,

1996;Facer

et al.,

1998; Nolano

et al.,

2000;Nolano

et al.,

2001) showed the reducedinnervation of the sweat glands in patientswith peripheral neuropathies, as well as inRoss’ syndrome, familial dysautonomia,and generalized anhidrosis.

Methods to perform skin biopsy and to quantify nerve fibres

Skin biopsy is most commonly performedby a disposable 3-mm punch, using sterile

technique and local anaesthesia with lido-caine. No suture is needed and no sideeffects have been reported. Healing is usuallycomplete within one week and the scar isbarely visible after 3 months. Punch biopsyallows obtaining a sample of skin thatincludes epidermis and superficial (sub-papillary and reticular) dermis.Two immunostaining methods are mostcommonly used for diagnostic purposes:bright-field immunohistochemistry and in-direct immunofluorescence with or withoutconfocal microscopy. Using bright-fieldmicroscopy, individual nerve fibres crossingthe dermal-epidermal junction are countedat high magnification in at least three sec-tions of 50

μ

m thickness. The length ofthe epidermis is measured with compute-rized software and the linear density ofintraepidermal nerve fibre

per

millimetre(IENF/mm) is therefore calculated.Indirect immunofluorescence study withconfocal laser microscopy is based onthree-dimension reconstruction of a varia-ble number of image stacks of sections cutat 80-100

μ

m thickness. Linear IENF den-sity is quantified using software for imageanalysis by tracing nerve fibres in three di-mensions. This technique is particularlyuseful in the study of receptors, glands,and vessels, and multiple staining usingdifferent markers.The less invasive “blister technique” isbased on the use of a suction capsule thatseparates the epidermis from the dermis atthe junction, with no bleeding and need oflocal anaesthesia (Kennedy

et al.,

1999).Epidermal innervation is examined in anarea wider than the surface of 3-mm ver-tical sections, but no information on der-mal innervation is provided. Though thistechnique deserves interest, it has a limitedusefulness in clinical practise. A compre-hensive review on methods and rules forIENF counting has been recently publishedin Dyck & Thomas textbook on peripheralneuropathies (Kennedy

et al.,

2005).

Applications of skin biopsy in peripheral neuropathies

The possibility to investigate small calibrenerve fibres, which are mainly involved inpainful neuropathies but cannot be exami-

1268

Rev Neurol (Paris) 2007 ; 163 : 12, 1266-1270

G. LAURIA

ned by routine neurophysiologic tests,prompted the use of skin biopsy in clinicalpractise. The choice of biopsy location isimportant to assess the pattern of skin ner-ve fibre loss. In peripheral neuropathies,biopsy is commonly performed in the dis-tal region of the leg, 10 cm above the la-teral malleolus, and in the upper lateral si-te of the thigh. A length-dependentpattern, characterized by a lower IENFdensity in the distal than the proximal siteof the lower limb, is typical of axonal po-lyneuropathies and reflects the dying-backof small nerve fibres. This pattern distin-guishes axonal neuropathies from sensoryneuronopathies due to the degeneration ofdorsal root ganglion neurons, in which awidespread loss of nerve fibres is com-monly observed (Sghirlanzoni

et al.,

2005).Quantification of morphological changesof IENF, such as axonal swellings, whichare a frequent finding in peripheral neuro-pathies, has been used to monitor the pro-gression of neuropathy in patients withdiabetes and HIV infection. Increaseddensity of IENF swellings correlated withimpaired heat-pain threshold, presence ofdiffuse degenerative changes in dermalnerves, and predicted the progression toovert neuropathy (Lauria

et al.,

2003;Herrmann

et al.,

2004b).In demyelinating neuropathies, which pre-dominantly affect large fibres, skin biopsyallowed demonstrating the involvement al-so of small fibres. This has been observedin chronic demyelinating inflammatory po-

lyradiculoneuropathies (Chiang

et al.,

2002) and in Guillain-Barré syndrome, inwhich a correlation between skin denerva-tion, dyasautonomia, and poorer outcome

was reported (Pan

et al.,

2003). In anti-myelin associated glycoprotein neuro-pathy, specific deposits of IgM in smallmyelinated fibres of hairy and glabrousskin were described (Lombardi

et al.,

2005). Finally, pathogenic morphologicalabnormalities of myelinated fibres andmyelin gene expression have been demons-trated in patients with Charcot-Marie-Tooth disease, suggesting that skin biopsymight replace nerve biopsy in the evalua-tion of inherited neuropathies (Li

et al.,

2005; Sabet

et al.,

2006).

Skin biopsy in small fibre neuropathies

Different normative range and cut-offvalues of IENF density in neuropathypatients have been reported using eitherbright-field immunohistochemistry orconfocal microscope technique. However,no systematic study comparing the twomethods has been carried out and onlystudies using bright-field technique havebeen specifically designed to assess normalvalues and diagnostic yield.Recently, the guidelines of the EuropeanFederation of Neurological Societies onthe use of skin biopsy in the diagnosis ofperipheral neuropathy have been published(Lauria

et al.,

2005a). Quantification ofIENF density proved to be the most reliabletool for the diagnosis of small fibre neuro-pathy in patients with normal electrophysio-logical examination, showing a sensitivityof 69-82 p.cent and a specificity of97 p.cent, which are not influenced by theimmunohistochemical technique. The dia-gnostic efficiency for diagnosing idiopa-thic or secondary (diabetic, cytotoxic, oramyloid) small fibre neuropathy was93 p.cent, with a positive predictive valueof 95 p.cent and a negative predictivevalue of 91 p.cent (Koskinen

et al.,

2005).Skin biopsy allows diagnosing small fibreneuropathy better than sural nerve biopsy.A large comparative study (Herrmann

etal.,

1999) showed that skin and nervebiopsy findings were concordant in73 p.cent of patients, but demonstratedthat decreased IENF density at the distalleg was the only indicator of small fibreneuropathy in 23 p.cent of patients. In pa-tients with small fibre neuropathy, IENFdensity is inversely correlated with warmthreshold assessed by quantitative sensorytesting, whereas correlation with heat-painand cooling thresholds as well as measuresof autonomic dysfunction needs is weaker(Lauria

et al.,

2005a). One study in patientswith painful neuropathy and normal suralnerve conduction showed a direct correla-tion between decrease of IENF densityand medial plantar sensory nerve actionpotential amplitude, suggesting that themost distal large myelinated nerve fibresmay be affected in small fibre neuropathyand that their recording might be more

sensitive than sural nerve conduction study(Herrmann

et al.,

2004a).Despite skin biopsy is mostly used to exa-mine patients with painful neuropathy, thecorrelation between IENF density and seve-rity of neuropathic pain remains unclear.In diabetic painful neuropathy, patientswith more severe pain had a lower IENFdensity in the distal leg than those withlower pain. However, within the group ofpatients with pain, pain intensity did notcorrelate with IENF density (Sorensen

etal.,

2006). One study in HIV neuropathyfound only a trend toward an inverse cor-relation between IENF density with painintensity (Herrmann

et al.,

2006) Patientscan experience persisting neuropathic paindespite complete skin denervation at thedistal legs, indicating that the generator ofpain is located more proximally. On theother hand, in congenital insensitivity topain with anhidrosis (hereditary sensoryneuropathy type IV) there is a lack of skinnerves. In patients with impaired glucosetolerance, lifestyle intervention and meta-bolic correction induced a partial recoveryof IENF density and a decrease of neuro-pathic pain (Smith

et al.,

2006).

Regeneration of skin nerve fibres

Skin biopsy is minimally invasive and canbe repeated to monitor the course of aneuropathy or the effect of neuroprotecti-ve interventions. Skin nerve fibres can re-generate either spontaneously, such as indiabetic truncal neuropathy (Lauria

et al.

,1998) and after capsaicin application (Si-mone

et al.

, 1998; Nolano

et al.

, 1999), orafter treatment in steroid responsive neuro-pathy (Nodera

et al.

, 2003). In all cases,skin reinnervation is followed by the reco-very of sensation.Follow-up biopsies can be easily perfor-med adjacent to the scar of the formerbiopsy and within the same peripheral nervedistribution. It has been recently demons-trated (Polydefkis

et al.

, 2004) that theregeneration rate of IENF is slower in dia-betic patients irrespective of the presenceof neuropathy or not, suggesting that dia-betes per se causes a functional impair-ment of peripheral axon re-growth. Thismethod could give important information

© 2007. Elsevier Masson SAS. Tous droits réservés

Actualités •

Skin biopsy

1269

G. LAURIA

in future clinical trials. In experimentalmodels of neuropathies (Bianchi

et al.

,2004; Bianchi

et al.

, 2006; Lauria

et al.

,2005b), IENF quantification proved tocorrelate with electrophysiological andbehavioural changes and is currently usedas an outcome measure in neuroprotectivetrials.

References

B

IANCHI

R, B

RINES

M, L

AURIA

G,

et al.

(2006).Protective effect of erythropoietin and of itscarbamylated derivative in experimental cis-platin peripheral neurotoxicity.

Clin CancerRes, 12:

2607-2612.B

IANCHI

R, B

UYUKAKILLI

B, B

RINES

M,

et al.

(2004). Erythropoietin both protects fromand reverses experimental diabetic neuropa-thy.

Proc Natl Acad Sci USA,

101:

823-828. C

HIANG

MC, L

IN

YH, P

AN

CL, T

SENG

TJ,L

IN

WM, H

SIEH

ST. (2002). Cutaneous inner-vation in chronic inflammatory demyelinatingpolyneuropathy.

Neurology, 59:

1094-1098.C

HIEN

HF, T

SENG

TJ, L

IN

WM,

et al.

(2001).Quantitative pathology of cutaneous nerveterminal degeneration in the human skin.

Acta

Neuropathol, 102:

455-461.D

ALSGAARD

CJ, R

YDH

M, H

AEGERSTRAND

A.(1989) Cutaneous innervation in man visua-lized with protein gene product 9.5 (PGP 9.5)antibodies.

Histochemistry, 92:

385-390.F

ACER

P, M

ATHUR

R, P

ANDYA

SS, L

ADIWALA

U,S

INGHAL

BS, A

NAND

P. (1998). Correlationof quantitative tests of nerve and targetorgan dysfunction with skin immunohisto-logy in leprosy.

Brain, 121:

2239-2247.G

ØRANSSON

LG, M

ELLGREN

SI, L

INDAL

S,O

MDAL

R. (2004). The effect of age andgender on epidermal nerve fiber density.

Neurology, 6:

774-777.H

ERRMANN

DN, G

RIFFIN

JW, H

AUER

P, C

ORN-

BLATH

DR, M

C

A

RTHUR

JC. (1999). Epider-mal nerve fiber density and sural nervemorphometry in peripheral neuropathies.

Neurology, 53:

1634-1640.H

ERRMANN

DN, F

ERGUSON

ML, P

ANNONI

V,B

ARBANO

RL, S

TANTON

M, L

OGIGIAN

EL.(2004a) Plantar nerve AP and skin biopsy insensory neuropathies with normal routineconduction studies.

Neurology,

63:

879-85.H

ERRMANN

DN, M

C

D

ERMOTT

MP, H

ENDER-

SON

D, C

HEN

L, A

KOWUAH

K, S

CHIFITTO

Gand The North East Aids Dementia (Nead)Consortium. (2004b). Epidermal nerve fiberdensity, axonal swellings and QST as pre-dictors of HIV distal sensory neuropathy.

Muscle Nerve, 29:

420-427.

H

ERRMANN

DN, MCDERMOTT MP, SOWDEN JE,

et al. (2006). Is skin biopsy a predictor of

transition to symptomatic HIV neuropathy?

A longitudinal study. Neurology, 66: 857-

861.

KARANTH SS, SPRINGALL DR, LUCAS S, et al.

(1989). Changes in nerves and neuropepti-

des in skin from 100 leprosy patients inves-

tigated by immunocytochemistry. J Pathol,

157: 15-26.

KENNEDY WR, NOLANO M, WENDELSCHAFER-

CRABB G, JOHNSON TL, TAMURA E. (1999).

A skin blister method to study epidermal

nerves in peripheral nerve disease. Muscle

Nerve, 22: 360-371.

KENNEDY WR, WENDELSCHAFER-CRABB G,

JOHNSON T. (1996). Quantitation of epider-

mal nerves in diabetic neuropathy. Neuro-

logy, 47: 1042-1048.

KENNEDY WR, WENDELSCHAFER-CRABB G,

POLYDEFKIS M, MCARTHUR J. (2005). Patho-

logy and quantitation of cutaneous nerves;

In: Dyck PJ, Thomas PK. (eds.), Peripheral

Neuropathy, 4th edition, Philadelphia, Saun-

ders; pp. 869-896.

KOSKINEN M, HIETAHARJU A, KYLÄNIEMI M, et

al. (2005). A quantitative method for the

assessment of intraepidermal nerve fibers in

small-fiber neuropathy. J Neurol, 252: 789-

794.

LAURIA G, BORGNA M, MORBIN M, et al.

(2004). Tubule and neurofilament immuno-

reactivity in human hairy skin: markers for

intraepidermal nerve fibers. Muscle Nerve,

30: 310-316.

LAURIA G, CORNBLATH DR, JOHANSSON O, et

al. (2005a). EFNS guidelines on the use of

skin biopsy in the diagnosis of peripheral

neuropathy. Eur J Neurol, 12: 1-12.

LAURIA G, HOLLAND N, HAUER PE, CORN-

BLATH DR, GRIFFIN JW, MCARTHUR JC.

(1999). Epidermal innervation: changes with

aging, topographic location, and in sensory

neuropathy. J Neurol Sci, 164: 172-178.

LAURIA G, LOMBARDI R, BORGNA M, et al.

(2005b). Intraepidermal nerve fiber density

in rat foot pad: neuropathologic-neurophysio-

logic correlation. J Periph Nervous Syst, 10:

202-208.

LAURIA G, MCARTHUR JC, HAUER PE, GRIF-

FIN JW, CORNBLATH DR. (1998). Neuro-

pathologic alterations in diabetic truncal

neuropathy: evaluation by skin biopsy. J Neu-

rol Neurosurg Psychiatry, 65: 762-766.

LAURIA G, MORBIN M, LOMBARDI R, et al.(2003). Axonal swellings predict the dege-neration of epidermal nerve fibers in painfulneuropathies. Neurology, 61: 631-636.

LAURIA G, MORBIN M, LOMBARDI R, et al.(2006). Expression of capsaicin receptorimmunoreactivity in human peripheral ner-vous system and in painful neuropathies. JPeriph Nerv Syst, 11: 262-271.

LI J, BAI Y, GHANDOUR K, QIN P, et al. (2005).Skin biopsies in myelin-related neuropa-thies: bringing molecular pathology to thebedside. Brain, 128: 1168-1177.

LI Y, HSIEH ST, CHIEN HF, ZHANG X, MCAR-

THUR JC, GRIFFIN JW. (1997). Sensory andmotor denervation influence epidermal thic-kness in rat foot glabrous skin. Exp Neurol,147: 452-462.

LOMBARDI R, ERNE B, LAURIA G, et al. (2005).Anti-MAG neuropathy patients show speci-fic IgM deposits in cutaneous nerve fibers.Ann Neurol, 57: 180-187.

MCARTHUR JC, STOCKS EA, HAUER P, CORN-

BLATH DR, GRIFFIN JW. (1998). Epidermalnerve fiber density: normative referencerange and diagnostic efficiency. Arch Neu-rol, 55: 1513-1520.

NODERA H, BARBANO RL, HENDERSON D,HERRMANN DN. (2003). Epidermal reinner-vation concomitant with symptomatic impro-vement in a sensory neuropathy. MuscleNerve, 2: 507-509.

NOLANO M, CRISCI C, SANTORO L, et al.(2000). Absent innervation of skin andsweat glands in congenital insensitivity topain with anhidrosis. Clin Neurophysiol,111: 1596-1601.

NOLANO M, PROVITERA V, CRISCI C, et al.(2001). Small fibers involvement in Frie-dreich’s ataxia. Ann Neurol, 50: 17-25.

NOLANO M, PROVITERA V, CRISCI C, et al.(2003). Quantification of myelinated endingsand mechanoreceptors in human digital skin.Ann Neurol, 54: 197-205.

NOLANO M, SIMONE DA, WENDELSCHAFER-CRABB G, JOHNSON TL, HAZEN E, KENNEDY

WR. (1999). Topical capsaicin in humans:parallel loss of epidermal nerve fibers andpain sensation. Pain, 81: 135-145.

PAN CL, LIN YH, LIN WM, TAI TY, HSIEH ST.(2001). Degeneration of nociceptive nerveterminals in human peripheral neuropathy.Neuroreport, 12: 787-792.

PAN CL, TSENG TJ, LIN YH, CHIANG MC,LIN WM, HSIEH ST. (2003). Cutaneous

1270 Rev Neurol (Paris) 2007 ; 163 : 12, 1266-1270

G. LAURIA

innervation in Guillain-Barré syndrome:pathology and clinical correlations. Brain,126: 386-397

POLYDEFKIS M, HAUER P, SHETH S, SIR-

DOFSKY M, GRIFFIN JW, MCARTHUR JC.(2004). The time course of epidermalnerve fibre regeneration: studies in normalcontrols and in people with diabetes, withand without neuropathy. Brain, 127:1606-1615.

SABET A, LI J, GHANDOUR K, PU Q, et al. (2006).Skin biopsies demonstrate MPZ splicingabnormalities in Charcot-Marie-Tooth neuro-pathy 1B. Neurology, 67: 1141-1146.

SGHIRLANZONI A, PAREYSON D, LAURIA G.(2005). Sensory neuron diseases. LancetNeurol, 4: 349-361.

SHUN CT, CHANG YC, WU HP, et al. (2004).Skin denervation in type 2 diabetes: correla-tions with diabetic duration and functionalimpairments. Brain, 127: 1593-1605.

SIMONE DA, NOLANO M, JOHNSON T, WEN-

DELSCHAFER-CRABB G, KENNEDY WR.(1998). Intradermal injection of capsaicin inhuman produces degeneration and subse-quent reinnervation of epidermal nervefibres: correlation with sensory function. JNeurosci, 18: 8947-8959.

SMITH AG, RUSSELL J, FELDMAN EL, et al.(2006). Lifestyle intervention for pre-diabe-tic neuropathy. Diabetes Care, 29: 1294-1299.

SORENSEN L, MOLYNEAUX L, YUE DK. (2006).The relationship among pain, sensory loss,and small nerve fibers in diabetes. DiabetesCare, 29: 883-887.

STÄNDER S, MOORMANN C, SCHUMACHER M,et al. (2004). Expression of vanilloidreceptor subtype 1 in cutaneous sensorynerve fibers, mast cells, and epithelialcells of appendage structures. Exp Derma-tol, 13: 129-139.