distal symmetric polyneuropathy · 2011. 2. 24. · oduct: pp08 ‐ urse descript istal symmetri...

J. D. EAsburyMD, L

No one in

The ide

Distangland, MD

y, MD, G. T. J. Kinsell

nvolved in the

America

eas and opinio

al SymD, G. S. GrT. Carter, Mla, MD, N.

e planning of Aut

CM

an Associatio

2621 S

ons in this Pra

mmetrronseth, MMD, J. A. CLatov, MD

Sumn

this CME actithors/faculty

ME is availabl

Copy

on of Neurom

Superior Dr N

actice Guidelirepresent th

ric PoMD, G. Fran

Cohen, MDD, R. A. Lewner, MD, K

ivity have anyhave nothing

le 8/31/2009

yright: 200

muscular and

NW Rochest

ine are solelyhose of the AA

olyneunklin, MD,D, M. A. Fiwis, MD, P

K. S

y relevant finag to disclose.

- 8/31/2012

05

d Electrodiag

ter, MN 559

y those of the ANEM

uropat, R. G. Millisher, MD, P. A. Low,

ancial relatio

gnostic Medi

01

author and do

Produc

thy ler, MD, AJ. F. HowaMD, and A

nships to disc

icine

o not necessa

ct: PP08

A. K. ard,

A. J.

close.

rily

CMPr CoDstteInTheldi LeU

ReExexD AcThCoboon HoOpuvi

ME Informatioroduct: PP08 ‐

ourse Descriptistal symmetritandardized, thesting, nerve bntended Audiehis course is ectrodiagnostisorders.

earning Objectpon conclusion

1. acknowreview

2. describ3. recogn

assessm

elease Date: 8xpiration Datexpiration date.uration/Comp

ccreditation anhe American Aouncil for Conoth sections ofnly the credit c

ow to Obtain Cnce you have urchased this pew a transcr

on Distal Symme

tion c polyneuropahe available litiopsy, and skinnce intended for ic medicine w

tives n of this prograwledge recommand analysis o

be the role of laize evidence‐bment of polyne

/31/2009 e: 8/31/2012. Y pletion Time: 3

nd DesignationAssociation of ntinuing Medicf this enduringcommensurate

CME reviewed the product. Answript of your

etric Polyneuro

thy (DSP) is thterature was rn biopsy for the

Neurologists,with the inten

am, participanmendations foof the peer‐revaboratory and based guidelineeuropathy.

Your request to

3 hours

n Statements Neuromuscul

cal Education tg material for ae with the exte

materials, youwer the questioCME by log

opathy

e most commoreviewed to pe assessment o

, Physiatrists, t to improve

ts should be abr the evaluatioiewed literatugenetic tests fes regarding th

o receive AMA

ar and Electroto provide cona combined mant of their part

u can obtain Cons and click sugging into ww

on variety of nrovide evidencof polyneuropa

and others wthe quality o

ble to: on of distal symre. for the assessmhe role of auto

PRA Category

odiagnostic Mntinuing medicaximum of 3 Aticipation in th

CME credit by ubmit. Once yww.aanem.org

europathy. Since‐based guideathy.

who practice of medical car

mmetric polyn

ment of polyneonomic testing

1 Credits™ mu

edicine (AANEcal education MA PRA Categhe activity.

clicking on thyour answers hg and clicking

nce the evaluatelines regardin

neuromusculre to patients

neuropathy(DS

europathy. g, nerve biopsy

ust be submitte

EM) is accredifor physiciansgory 1 creditsTM

e link in the ehave been submg View Profi

tion of this disng the role of

ar, musculosks with muscle

SP) based on a

y, and skin bio

ed on or befor

ited by the Acs. The AANEMM. Physicians s

e‐mail receivedmitted, you wile and then

order is not f autonomic

keletal, and and nerve

a prescribed

opsy for the

re the credit

ccreditation designates hould claim

d when you ll be able to My CME.

AAEM PRACTICE TOPIC IN ELECTRODIAGNOSTIC MEDICINEAmerican Association of Electrodiagnostic Medicine421 First Avenue S.W., Suite 300 East, Rochester, MN 55902 (507/288-0100)

ABSTRACT: The objective of this report was to develop a case definitionof “distal symmetrical polyneuropathy” to standardize and facilitate clinicalresearch and epidemiological studies. A formalized consensus process wasemployed to reach agreement after a systematic review and classification ofevidence from the literature. The literature indicates that symptoms alonehave relatively poor diagnostic accuracy in predicting the presence of poly-neuropathy; signs are better predictors of polyneuropathy than symptoms;and single abnormalities on examination are less sensitive than multipleabnormalities in predicting the presence of polyneuropathy. The combina-tion of neuropathic symptoms, signs, and electrodiagnostic findings providesthe most accurate diagnosis of distal symmetrical polyneuropathy. A set ofcase definitions was rank ordered by likelihood of disease. The highestlikelihood of polyneuropathy (useful for clinical trials) occurs with a combi-nation of multiple symptoms, multiple signs, and abnormal electrodiagnosticstudies. A modest likelihood of polyneuropathy (useful for field or epidemi-ological studies) occurs with a combination of multiple symptoms and mul-tiple signs when the results of electrodiagnostic studies are not available. Alower likelihood of polyneuropathy occurs when electrodiagnostic studiesand signs are discordant. For research purposes, the best approach fordefining distal symmetrical polyneuropathy is a set of case definitions rankordered by estimated likelihood of disease. The inclusion of this formalizedcase definition in clinical and epidemiological research studies will ensuregreater consistency of case selection.

Muscle Nerve 31: 113–123, 2005

DISTAL SYMMETRICAL POLYNEUROPATHY:DEFINITION FOR CLINICAL RESEARCH

J. D. ENGLAND, MD, G. S. GRONSETH, MD, G. FRANKLIN, MD, R. G. MILLER, MD, A. K. ASBURY, MD,

G. T. CARTER, MD, J. A. COHEN, MD, M. A. FISHER, MD, J. F. HOWARD, MD, L. J. KINSELLA, MD,

N. LATOV, MD, R. A. LEWIS, MD, P. A. LOW, MD, and A. J. SUMNER, MD

American Association of Electrodiagnostic Medicine, 421 First Avenue SW,Suite 300E, Rochester, MN 55902, USA

The American Association of ElectrodiagnosticMedicine (AAEM) in conjunction with the AmericanAcademy of Neurology (AAN) and the AmericanAcademy of Physical Medicine and Rehabilitation(AAPM&R) determined that there was a need for aformal case definition of polyneuropathy. Because of

inconsistency in the literature, no consistent casedefinition exists. The use of a formal case definitionacross future research studies would ensure greaterconsistency of patient selection. This review de-scribes the development of such a case definition for“distal symmetrical polyneuropathy.”

This article was prepared and reviewed by the AAEM and did notundergo the separate review process of Muscle & Nerve.

This article is a joint report of the American Association of ElectrodiagnosticMedicine, the American Academy of Neurology, and the American Academyof Physical Medicine and RehabilitationAbbreviations: EMG, electromyography; MDNS, Michigan diabetic neurop-athy score; MNSI, Michigan neuropathy screening instrument; NCS, nerve-conductions study/studies; NDS, neuropathy disability score; NIS-LL, neu-ropathy impairment score in the lower limbs; QST, quantitative sensory testing

Key words: case definition; clinical research; electrodiagnosis; epidemiol-ogy; polyneuropathyCorrespondence to: T. Schmidt; e-mail: [email protected]

© 2004 American Association of Electrodiagnostic Medicine. Published byWiley Periodicals, Inc.Published online 9 November 2004 in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/mus.20233

Distal Symmetrical Polyneuropathy MUSCLE & NERVE January 2005 113

Polyneuropathy is a common neurological disor-der with a diverse etiology. Although experiencedclinicians can usually diagnose polyneuropathy inpatients presenting with the characteristic historyand classic neurological examination findings, theexact criteria for diagnosis are not formalized. Inparticular, accurate criteria for the diagnosis of “dis-tal symmetrical polyneuropathy” are debated.

The principal purpose of this project was to de-velop a definition of distal symmetrical polyneurop-athy with a reasonably high sensitivity and specificitythat would serve as a basis for future research studies.Clinicians may find the criteria useful for routineclinical diagnosis. To achieve greater focus, otherneuropathy phenotypes, including polyradiculopa-thy, mononeuropathy multiplex, Guillain–Barré syn-drome, chronic inflammatory demyelinating poly-neuropathy, and related conditions, were excludedfrom the final case definition. Although “small-fiber”polyneuropathy is an important subset of “distal sym-metrical polyneuropathy,” the evidence-based medi-cal literature is insufficient to provide an adequatecase definition for isolated or pure small-fiber poly-neuropathy at this time.

The case definition of “distal symmetrical poly-neuropathy” described herein is based on a system-atic analysis of peer-reviewed literature supple-mented by consensus from an expert panel.

PROCESS

Formation of Expert Panel. The polyneuropathy taskforce included 14 physicians with representativesfrom the AAN, AAEM, and AAPM&R. All task forcemembers had extensive experience and expertise inthe area of polyneuropathy. In addition, three phy-sicians with expertise in evidence-based methodol-ogy and practice parameter development partici-pated in the project.

Finding the Best Evidence. The literature search in-cluded OVID Medline (1970 to April 2004), OVIDExcerpta Medica (EMBASE; 1980 to April 2004),and OVID Current Contents (2000 to April 2004).The search included articles on humans only and inall languages. The search terms selected were poly-neuropathy, distal symmetrical polyneuropathy, dis-tal axonopathy, fiber length–dependent polyneu-ropathy, and distal axonal loss polyneuropathy. Thesearch terms, mononeuropathy, mononeuropathymultiplex, radiculopathy, polyradiculopathy, plex-opathy, multifocal motor neuropathy, acute inflam-matory demyelinating polyneuropathy, Guillain–Barré syndrome, and chronic inflammatory de-

myelinating polyneuropathy, were included onlywhen they appeared in studies wherein the primaryfocus was “distal symmetrical polyneuropathy.”

Panel experts were asked to identify additionalarticles missed by the initial search strategy. Further-more, the bibliographies of the selected articles werereviewed for potentially relevant articles.

Three committee members reviewed the titlesand abstracts of citations identified from this origi-nal search for those that were potentially relevant fordefining “distal symmetrical polyneuropathy.” Arti-cles considered potentially relevant by any panelmember were also obtained.

Potentially relevant articles were subsequently re-viewed in their entirety by three reviewers and wereincluded in the initial analysis if they met the follow-ing criteria: (1) The study included patients with andwithout distal symmetrical polyneuropathy. In orderto assess the likelihood of “spectrum bias,” the char-acteristics of the comparison group without distalsymmetrical polyneuropathy were noted. Thosestudies in which the control group included subjectswith neuropathic features that may mimic or overlapwith “distal symmetrical polyneuropathy” were ratedas more relevant. (2) The patients had a potentialdiagnostic predictor (i.e., symptom, sign, or test re-sult) measured. (3) The patients were determined tohave a distal symmetrical polyneuropathy on thebasis of an explicitly defined independent referencestandard (an acceptable standard was not prespeci-fied by panel members). (4) The presentation of thedata in the study allowed calculation of sensitivitiesand specificities.

From each study the following methodologicalcharacteristics were abstracted (see Appendix 1,Glossary of Terms): the study design (case–control,cross-sectional, cohort survey); the number of pa-tients; the target disorder, including the spectrum ofseverity of the target disorder; the diagnostic predic-tor(s); the reference standard employed; whetherthe reference standard was measured without knowl-edge of the result of the diagnostic predictor; theproportion of patients with the target disorder whowere positive for the diagnostic predictor (sensitiv-ity); and the proportion of patients without the tar-get disorder who were negative for the diagnosticpredictor (specificity).

Each reviewer graded the risk of bias in eachstudy by using the diagnostic test classification-of-evidence scheme in Appendix 2. In this scheme,articles attaining a grade of Class IV are judged tohave the highest risk of bias, and articles attainingClass I are judged to have the lowest risk of bias. Onlystudies attaining a grade of Class I, II, or III were

114 Distal Symmetrical Polyneuropathy MUSCLE & NERVE January 2005

further considered in the analysis. In the grading ofstudies, electrodiagnostic studies were considered an“objective” outcome. Disagreements among the re-viewers regarding an article’s grade were resolvedthrough discussion.

Consensus Process. A formal consensus process(nominal group process)12,22 was used to develop thecase definition. Because there is no single “goldstandard” that defines distal symmetrical polyneu-ropathy, the case definition must account for differ-ent levels of certainty for the presence or absence ofthe disorder. In line with this goal, participants weregiven several guidelines for developing a case defi-nition. The case definition should: (1) be restrictedto “distal symmetrical polyneuropathy”; (2) serve as adefinition for the identification of cases in researchstudies; (3) acknowledge varying levels of diagnosticcertainty by including a set of case definitions rankordered by estimated ordinal likelihood of disease;(4) be simple, practical, and widely applicable bypracticing clinicians; and (5) be based, as much aspossible, on current best evidence.

Through several face-to-face meetings, electronicmail, and telephone conferences, committee mem-bers reviewed the results of the literature review andproposed case definitions of varying ordinal likeli-hood of distal symmetrical polyneuropathy. Points ofagreement and disagreement were identified, dis-cussed, and resolved. The elements of the proposed

case definitions were repeatedly tested against theconclusions from the literature review. What evolvedfrom this process was an ordered set of case defini-tions ranked by likelihood of disease. The essence ofthe case definition procedure is shown in Tables 1and 2.

The Quality Standards Subcommittee of theAAN, the Practice Issues Review Panel of the AAEM,and the Practice Guidelines Committee of theAAPM&R (Appendix 3), reviewed and approved adraft of this paper with the proposed case definition.The draft was then sent to members of the AAN,AAEM, and AAPM&R for further review and then tothe journal Neurology for peer review. Boards of theAAN, AAEM, and AAPM&R reviewed and approvedthe final version of the paper. At each step of thereview process, external reviewers’ suggestions wereexplicitly considered. When appropriate, the expertpanel made changes to the document.

EVIDENCE

The search yielded 1450 references. After reviewingtitles and abstracts, 61 articles were retrieved andreviewed in their entirety. After comprehensive re-view of these papers, 12 articles attained a grade ofClass I, II, or III.2,3,7–10,15,17,18,24–26 These articlesserve as the major evidence basis for the case defini-tion and are tabulated in Table 3.

Table 1. Estimated likelihood of distal symmetrical polyneuropathy for case definitions that include symptoms, signs, and nerve conductionstudies (recommendations for clinical research studies).

Neuropathicsymptoms

Decreased or absentankle reflexes*

Decreased distalsensation

Distal muscle weaknessor atrophy NCS†

Ordinallikelihood

Present Present Present Present Abnormal ��Absent Present Present Present Abnormal ��Present Present Present Absent Abnormal ��Present Present Absent Absent Abnormal ��Present Absent Present Absent Abnormal ��Absent Present Absent Present Abnormal ��Present Absent Absent Absent Abnormal ��Absent Absent Absent Absent Abnormal ��Absent Present Absent Absent Abnormal ��Present Present Present Absent Normal ��Present‡ Absent Present‡ Absent Normal‡ �Present§ Present§ Present§ Present§ Normal§ �

Neuropathic symptoms: numbness, altered sensation, or pain in the feet. NCS, nerve conduction studies. For clinical research studies enrollment should belimited to cases above the bold horizontal line (i.e., ��).*Ankle reflexes may be decreased in normal individuals �65–70 years.†Abnormal NCS is defined in text.‡This phenotype is common in “small-fiber” sensory polyneuropathy. Determination of intraepithelial nerve fiber density in skin biopsy may be useful to confirmthe diagnosis (see text).§This phenotype in the presence of normal NCS is not a distal symmetrical polyneuropathy. This situation is given a negative (�) ordinal likelihood because thecondition cannot be classified as a distal symmetrical polyneuropathy. It is included here to emphasize the importance of including NCS as part of the casedefinition for clinical research studies.


Study Characteristics. Diabetic peripheral neurop-athy, which is the most prevalent and rigorouslystudied type of distal symmetrical polyneuropathy,was the target disorder in most studies. There is arelative lack of high-quality evidence for other vari-eties of distal symmetrical polyneuropathy. However,three of the studies (25% of the total) focused oncryptogenic sensory peripheral neuropathy. Al-though limited in quantity, the quality of the articleswas high and allowed the development of a casedefinition for “distal symmetrical polyneuropathy.”

The diagnostic predictors studied varied. Severalarticles described the diagnostic accuracy of singlesymptoms including foot numbness, foot pain, andcomplaints of “sensory alteration.” In addition, somearticles measured the accuracy of more complexcomposite symptom checklists. The accuracy of sin-gle examination elements was also determined.These included absent ankle reflexes, decreased dis-tal lower extremity strength, and decreased vibrationor cold detection. Some articles also measured theaccuracy of composite examinations that includedtwo or more examination elements.

The studies used different reference standards todetermine the presence of a symmetric distal periph-eral neuropathy. These included nerve conductionstudies (NCS), a clinician’s global impression, andcomposite clinical examination scores.

All studies collected data prospectively. Mostwere cohort surveys, but some used a case–controldesign. Four studies described measuring the pres-ence of a polyneuropathy using the reference stan-dard in a fashion that was masked to measurement ofthe diagnostic predictor. Two studies attained agrade of Class I,8,9 five attained a grade of ClassII,2,10,15,18,25 and five attained a grade of ClassIII.3,7,17,24,26

Diagnostic Accuracy. The diagnostic accuracy ofthe predictors was determined by calculating theirsensitivities and specificities. One way of displayingthese data is to plot sensitivities against specificitiesin a receiver operator characteristics (ROC) curve(Fig. 1).

Predictors encompassing a single specific symp-tom such as foot numbness have low sensitivity buthigh specificity for the presence of polyneuropathy.Predictors incorporating the presence of any one ofa number of neuropathic symptoms, such as thepresence of foot numbness or pain, attain a greatersensitivity but have lower specificity.

Particular single examination findings, such asabsent ankle tendon reflexes, have moderate sensi-tivity and high specificity for the presence of poly-neuropathy. When individual examination findingsare combined into a composite examination score,higher diagnostic accuracy results. The examinationscores with the highest sensitivity and specificity in-clude the screening examination used in the SanLuis Valley Diabetes Study,8 the neuropathy disabil-ity score (NDS),2–6 the neuropathy impairmentscore in the lower limbs (NIS-LL),3 the Michiganneuropathy screening instrument (MNSI) the Mich-igan diabetic neuropathy score (MDNS),7 and twoother well-described clinical examination scores.17,25

Notably, simple composite examination scores are asaccurate as more complex examinations.

The sensitivities and specificities of quantitativesensory testing (QST) varied widely among studies.These psychophysical tests have greater inherentvariability, making their results more difficult to stan-dardize and reproduce. Reproducibility of QST var-ied from poor to excellent.21,23 For these reasons,QST was not included as part of the final case defi-nition.

Table 2. Estimated likelihood of distal symmetrical polyneuropathy for case definitions that include only symptoms and signs(recommendations for field or epidemiological studies).

Neuropathicsymptoms

Decreased or absentankle reflexes*

Decreased distalsensation

Distal muscle weaknessor atrophy NCS†

Ordinallikelihood

Present Present Present Present ND ��Present Present Present Absent ND ��Present‡ Absent Present‡ Absent ND �Present Absent Absent Absent ND �Absent Present Absent Absent ND �

Neuropathic symptoms: numbness, altered sensation, or pain in the feet NCS, nerve conduction studies. For field epidemiology studies enrollment should belimited to cases above the bold horizontal line (i.e., ��).*Ankle reflexes may be decreased in normal individuals �65–70 years.†Nerve conduction studies (NCS) are not included as part of the case definitions for epidemiology studies. ND, not done.‡This phenotype is common in “small-fiber” sensory polyneuropathy. Determination of intraepithelial nerve fiber density in skin biopsy may be useful to confirmthe diagnosis (see text).


Table 3. Studies meeting inclusion criteria.

Article(referencenumber) Target disorder Predictor

Referencestandard Cases Controls Design Spectrum Masked Class

Sensitivity(%)

Specificity(%)

9 Diabetic PN Symptom checklist“pain,” “sensoryalteration,” “feetnumbness”

Clinical examscore �4

188 400 Ch B Y 1 18 91

26 9128 93

8 Diabetic PN 2 of 3�:symptoms, abntemp. sens,2ankle DTRs

Neurologistclinicalevaluation

15 23 Ch B Y 1 87 91

10 Diabetic PN Symptomquestionnaire,neurologicexam, vibrationdetection

NCS 47 157 Ch N ND 2 87 60

94 9264 9718

18 Chronicsymmetric PNin elderly

Neuropathysymptoms

Bilateralimpairedsensation,strength, orDTR

11 9 CC B Y 2 78 82

2 Diabeticneuropathy

Symptom score,disability score,vibrationdetection, colddetection

Two or moreabn NCS

125 55 Ch N ND 2 70 84

65 9159 8644 87

15 Diabetic PN Vibration detectionthreshold,thermalthreshold

Clinically overtneuropathy

Ch N Y 2 100 43

43 7617 Diabetic PN Neuropathy exam Monofilaments

vibrationdetection

23 50 Ch B ND 3 96 51

26 CIAP vs. CIDP Absent ankleDTRs, � bicepsand � ankleDTRs

Publishedcriteria

11 11 CC N ND 3 100 18

100 9124 CIAN vs. HSMN Onset sensory,

onset motor,absent ankleDTR

Family history 48 47 CC N ND 3 60 85

40 1575 11

3 Diabeticpolyneuropathy

NIS-LL NIS-LL � 7tests

58 137 Ch B ND 3 69 87

Abn ankle DTR,Abn vibration,One or moreabn NCS, Twoor more abnNCS

60 91

17 9693 5881 91


Exam scoringsystem

NCS 49 29 Ch N ND 2 88 83


Symptoms,sensory exam,strength exam,reflexes,compositeexam, screeningexam

Mayo criteria Ch B ND 3 74 55

74 10059 10080 10080 10080 95

Abn, abnormal; B, broad spectrum of patients included; CC, case control; Ch, cohort survey; CIAN, chronic idiopathic ataxic polyneuropathy; CIAP, chronic idiopathic axonalneuropathy; CIDP, chronic inflammatory demyelinating polyneuropathy; DTRs, deep tendon reflexes; HSMN, hereditary sensory motor neuropathy; LL, lower limb; N, narrowspectrum of patients included; NCS, nerve conduction studies; ND, not described; NIS, neuropathy impairment score; PN, peripheral neuropathy; Sens, sensitivity; Spec,specificity; temp, temperature; Y, yes; �, positive; �, negative;2, decreased.


The sensitivities and specificities of quantitativeautonomic testing are relatively high for document-ing the presence or absence of autonomic dysfunc-tion.3,4 However, these tests are not routinely per-formed at all medical centers. Because a usable casedefinition must be based on tests that are simple,practical, and easily available, quantitative auto-nomic testing is not included as part of the final casedefinition.

Evidence-Based Conclusions for the Case Definition.

Using the definitions for strength of recommenda-tion (Appendix 4) the following conclusions andrecommendations can be supported from formalanalysis and classification of the literature:

1. Symptoms alone have relatively poor diagnosticaccuracy in predicting the presence of polyneu-ropathy. Multiple neuropathic symptoms are

more accurate than single symptoms and shouldbe weighted more heavily (Level B).

2. Signs are better predictors of polyneuropathythan symptoms and should be weighted moreheavily (Level B).

3. A single abnormality examination is less sensitivethan multiple abnormalities in predicting thepresence of polyneuropathy; therefore, an exam-ination for polyneuropathy should look for acombination of signs (Level B).

4. Relatively simple examinations are as accurate indiagnosing polyneuropathy as complex scoringsystems; therefore, the case definition can usesimple examinations without compromising accu-racy (Level B).

5. There is too much inconsistency among the stud-ies describing the accuracy of QST for its incor-poration into the case definition (Level U).

FIGURE 1. The diagnostic accuracy levels of symptoms, signs, or combinations of symptoms or signs (predictors) for the presenceof distal symmetric polyneuropathy are expressed. Predictors are plotted according to their sensitivity and specificity. Points plottednear the top of the graph correspond to predictors with high sensitivity for distal symmetric polyneuropathy. Points plotted near theleft side of the graph correspond to predictors with high specificity. Thus, points nearest the upper left-hand corner correspond topredictors with the highest diagnostic accuracy (both high sensitivity and specificity) for distal symmetric polyneuropathy. Pointsfalling near the diagonal line correspond to predictors with low diagnostic accuracy. Diamonds: diagnostic accuracy of symptoms;triangles: signs; shaded � or X symbols: quantitative sensory tests. Points describing the diagnostic accuracy of a single symptom(e.g., “numbness”) or a single examination finding (e.g., absent ankle reflexes) are enclosed by dashed ovals and circles. Pointsdescribing the diagnostic accuracy of more than one symptom (e.g., “numbness” or “pain”) or more than one sign (e.g., “absentankle reflexes” or “decreased distal sensation”) are not enclosed in dashed ovals and circles. The number just to the upper right ofeach plotted point indicates the study (reference no.) from which the sensitivity and specificity of that predictor was obtained.


CONSENSUS-BASED PRINCIPLES

The concept of distal symmetrical polyneuropathyrequires a clear definition of “distal” and “symmetri-cal” in the context of polyneuropathy. Distal refers tothose parts most distant from the center of the body.The polyneuropathy must begin in the feet. “Sym-metrical” indicates that the symptoms and signs arethe same on both sides of the body. Persistent orstriking asymmetry of symptoms or signs is inconsis-tent with the case definition. The case definitionmust encompass a description of symptoms and signswith an easily recognizable phenotype.

Symptoms. Symptoms may be primarily sensory,primarily motor, or both.3,7–9,17,18,24 Symptoms begindistally in the feet. Sensory symptoms are eitherpersistent or intermittent alterations of sensationinitially involving the toes or feet. Occasionally, anisolated digital sensory neuropathy affecting one ormore toes may be difficult to distinguish from anearly polyneuropathy. The differentiation may bediscernible only with time. Frequently described sen-sory symptoms include numbness, burning, prick-ling paresthesias, dysesthesias, and allodynia. Whenmotor symptoms are the first manifestation of poly-neuropathy, the patient may note weakness in thedistal legs. Distal symmetrical polyneuropathy maybe asymptomatic, especially in its early stage. Anasymptomatic presentation is more likely when pos-itive sensory symptoms, such as dysesthesias or par-esthesias, are lacking, or when motor deficits aloneare the presenting features. A number of symptomquestionnaires and methods for scoring symptomshave been described.2,3,7–10,15,17,18,24–26

Signs. Signs of distal symmetrical polyneuropathyevident on clinical examination may include abnor-malities of primary sensory modalities (pain, touch,hot, cold, vibration, and proprioception), the motorsystem (weakness and atrophy), tendon reflexes (es-pecially depressed or absent ankle jerks), or theautonomic system.

Signs of sensory loss occur in an acral, non-dermatomal, nonsingle-nerve distribution. Sensorysymptoms and their concomitant signs evolve in acentripetal manner.

Motor signs may include atrophy and weakness ofintrinsic foot muscles and associated foot deformi-ties such as hammer toes and pes cavus. Because pescavus does not always indicate a polyneuropathy, italone is not sufficient evidence of polyneuropathy.With centripetal progression of motor involvement,

weakness of toe dorsiflexion followed by weakness offoot dorsiflexion can be expected.

Tendon reflexes are often depressed or unelicit-able. Ankle jerks that are relatively depressed orunelicitable are valuable signs of polyneuropathy;however, the interpretation of such findings requiresconsiderable clinical experience and judgment. Inaddition, other possible causes of depressed or ab-sent ankle jerks, such as S-1 radiculopathy, focalneuropathies, and age-related decreases, must beexcluded.

Signs of autonomic nervous system involvementmay also constitute findings consistent with a distalsymmetrical polyneuropathy if small fibers are af-fected. Autonomic dysfunction should begin distallyand may include abnormalities of sweating or circu-latory instability in the feet.

Electrodiagnostic Studies. No single “referencestandard” defines distal symmetrical polyneurop-athy. The most accurate diagnosis of distal symmet-rical polyneuropathy comprises a combination ofclinical symptoms, signs, and electrodiagnostic find-ings. Electrodiagnostic findings should be includedas part of the case definition because they provide ahigher level of specificity for the diagnosis.3,5,7,24

Electrodiagnostic studies are sensitive, specific,validated measures of the presence of polyneurop-athy.2,3,4,5,7,10,19,20,24 Electrodiagnostic evaluationscommonly include both NCS and needle electro-myography (EMG). In the diagnosis of polyneurop-athy, NCS are the most informative part of theelectrodiagnostic evaluation.4,5,7,10,19,20,24 NCS arenoninvasive, standardized, and provide a sensitivemeasure of the functional status of sensory and mo-tor nerve fibers. NCS are also widely performed andsuitable for population studies or longitudinal eval-uations. The inclusion of NCS in the assessment ofpolyneuropathy adds a higher level of specificity tothe diagnosis.3,5,7,24 For these reasons, NCS are in-cluded as an integral part of the case definition ofpolyneuropathy.

The protocol for performing NCS was deter-mined by the structured consensus process de-scribed previously. There are have been many rec-ommendations regarding NCS criteria for thediagnosis of polyneuropathy, but no formal consen-sus exists. The recommendations that follow arebased on electrophysiological principles that com-bine both the highest sensitivity and specificity aswell as the highest efficiency for the diagnosis ofdistal symmetrical polyneuropathy.


Recommended Protocol for Nerve Conduction Studies.

The following set of sensory and motor NCS should beperformed if patients are entering a clinical researchtrial in which NCS will be tracked longitudinally. Thisprotocol includes unilateral studies of sural sensory,ulnar sensory, and median sensory nerves, and pero-neal, tibial, median, and ulnar motor nerves with Fwaves. Other NCS may be necessary as determined byclinical judgment. The minimum case definition crite-rion for electrodiagnostic confirmation of distal sym-metrical polyneuropathy is an abnormality (�99th or�1st percentile) of any attribute of nerve conductionin two separate nerves, one of which must be the suralnerve. Electrodiagnostic studies should follow rigorousguidelines such as those set by the AAEM.1 Variablessuch as skin temperature, age, height, gender, andweight should be measured and accounted for whenreporting a NCS as normal or abnormal.1

A simplified NCS protocol may be used for thepurpose of defining the presence of distal symmet-rical polyneuropathy. However, the abbreviated pro-tocol is not sufficient to determine the subtype orseverity of the polyneuropathy. For these purposes,as well as for clinical trials in which electrodiagnosticmeasures will be tracked serially, the more compre-hensive set of NCS is recommended.

The simplified NCS protocol is as follows:

1. Sural sensory and peroneal motor NCS are per-formed in one lower extremity. Taken together,these NCS are the most sensitive for detecting adistal symmetrical polyneuropathy. If both studiesare normal, there is no evidence of typical distalsymmetrical polyneuropathy. In such a situation,no further NCS are necessary.

2. If sural sensory or peroneal motor NCS are abnor-mal, then additional NCS are recommended. Thisshould include NCS of at least the ulnar sensory,median sensory, and ulnar motor nerves in oneupper extremity. A contralateral sural sensory andone tibial motor NCS may also be performed ac-cording to the discretion of the examiner. Cautionis warranted when interpreting median and ulnarstudies because there is a possibility of abnormalitydue to compression of these nerves at the wrist orulnar neuropathy at the elbow.

3. If a response is absent for any of the nerves stud-ied (sensory or motor), NCS of the contralateralnerve should be performed.

4. If a peroneal motor response is absent, an ipsilat-eral tibial motor NCS should be performed.

Minimal criteria for the electrodiagnostic confirma-tion of distal symmetrical polyneuropathy are thesame as those listed previously.

COMBINING EVIDENCE AND CONSENSUS: CASEDEFINITION OF DISTAL SYMMETRICALPOLYNEUROPATHY

The best approach for defining distal symmetricalpolyneuropathy is an ordered set of definitionsranked by likelihood of disease. The likelihood ofdistal symmetrical polyneuropathy was rated on anordinal scale from highest likelihood (��) tolowest likelihood (�). Because diagnostic certaintyfor polyneuropathy follows a continuum of probabil-ity, this manner of definition is the most sensible. Ineach set of case definitions a hierarchy of parametercombinations was established to provide the mostrelevant combinations for the diagnosis of distal sym-metrical polyneuropathy. Combinations of parame-ters that were considered clinically unusual and notappropriate for research studies were not included.For these reasons not every possible combination ofparameters is presented.

The essential characteristics of the case defini-tion are given in Tables 1 and 2. Important aspects ofthe case definition that warrant emphasis include:

1. The combination of neuropathic symptoms,signs, and abnormal electrodiagnostic studiesprovides the most accurate diagnosis of distal sym-metrical polyneuropathy (Table 1).

2. Electrodiagnostic studies are recommended aspart of the clinical research case definition (Table1) because they are objective and validated testsof peripheral nerve function. Abnormal electro-diagnostic studies increase the likelihood of thepresence of distal symmetrical polyneuropathyand provide a higher level of specificity to thecase definition. Electrodiagnostic studies shouldnot be used alone to make the diagnosis becausetheir sensitivity and specificity are not perfect.

3. Electrodiagnostic studies are not required forfield or epidemiological studies (Table 2), but thelikelihood of diagnosis must be downgraded ac-cordingly.

4. For research studies, enrollment should be lim-ited to cases that are most likely to have distalsymmetrical polyneuropathy (i.e., those thatachieve the highest specificity for the diagnosis).For clinical research studies, this consists of caseswith an ordinal likelihood of �� (Table 1).For epidemiological studies, this consists of caseswith an ordinal likelihood of �� (Table 2).

LIMITATIONS AND FUTURE RESEARCH

This case definition is heavily weighted toward distalsymmetrical polyneuropathy with predominant in-


volvement of “large fibers,” and it is not intended toemphasize the subset of distal symmetrical polyneu-ropathy termed small-fiber polyneuropathy. Becausethis type of polyneuropathy may present with onlypain and numbness in the feet accompanied by fewsigns and normal NCS, a formal case definition re-stricted to small-fiber polyneuropathy is difficult todevelop at this time. This is especially true becausethere is no widely available method to confirm thediagnosis. Determination of intraepithelial nerve fi-ber density in punch biopsies of skin is a promisingtechnique.11,13,14,16 Inclusion of small-fiber polyneu-ropathy in a formal case definition must await fur-ther studies.

Another limitation of the case definition is thatmost of the available best evidence is restricted todiabetic peripheral neuropathy. The reason that di-abetic neuropathy figures so prominently in theanalysis is that it is the most common and rigorouslystudied variety of distal symmetrical polyneuropathy.The other studies included in the analysis focusedon cryptogenic sensory peripheral neuropathy.Thus, some uncertainty exists with respect to thegeneralization of the case definition for distal sym-metical polyneuropathy associated with other etiolo-gies.

The process just described is an attempt todevelop formal criteria for a case definition ofdistal symmetrical polyneuropathy. The principalpurpose of the case definition is the identificationof cases for clinical research and epidemiologicalstudies. The criteria were formulated using a nom-inal group process in addition to the best availablescientific evidence. Validation and refinement ofthese criteria in future studies is encouraged. Spe-cifically, additional studies are needed before con-clusions can be made regarding the role of QSTand skin biopsy in the diagnosis of distal symmet-rical polyneuropathy. As quantitative autonomictesting becomes more routinely available, thesetests could easily be incorporated into the casedefinition. Future studies should also compare thecriteria delineated in this study with evolving, newcriteria. A major aim of the AAN, AAEM, andAAPM&R is that the case definition be modifiedand refined as new evidence accumulates.

APPENDIX 1: GLOSSARY OF TERMS

Predictor (diagnostic predictor): A symptom, exam-ination finding, or test result potentially predictingthe presence of a distal symmetrical polyneuropathy.

Target disorder: The condition or disease beingsought. In the current context, the target disorder

was a specific type of distal symmetrical polyneurop-athy (e.g., diabetic peripheral neuropathy).

Reference standard (“gold standard”): The testor procedure (or series of tests or procedures) per-formed to determine the actual presence or absenceof a distal symmetrical polyneuropathy.

Nominal group process: A formalized, iterativemethod for achieving consensus from a group ofexperts that attempts to maximize group reasoningwhile preserving individual input.

ROC (receiver operator characteristic) curve: Astandardized graph of sensitivity (true positive rate)by specificity (true negative rate) designed to depictdiagnostic accuracy and the trade-off between in-creasing sensitivity and decreasing specificity.

APPENDIX 2: DEFINITIONS FOR STRENGTH OFEVIDENCE

Diagnostic Evidence. Class I: Evidence provided by aprospective study of a broad spectrum of personswith the suspected condition. The study measuresthe diagnostic accuracy of the test using an accept-able independent reference standard for case defi-nition. The test, if not objective, is applied in anevaluation that is masked to the person’s clinicalpresentations and the reference standard is appliedin an evaluation that is masked to the test result.

Class II: Evidence provided by a prospective studyof a narrow spectrum of persons with the suspectedcondition, or by a retrospective study of a broadspectrum of persons with the condition comparedwith a broad spectrum of control subjects. The studymeasures the diagnostic accuracy of the test using anacceptable independent reference standard for casedefinition. The test is applied in an evaluation that ismasked to the reference standard.

Class III: Evidence provided by a retrospectivestudy when either the persons with the condition orthe control subjects are of a narrow spectrum. Thestudy measures the diagnostic accuracy of the testusing an acceptable independent reference standardfor case definition.

Class IV: Evidence provided by expert opinion orcase series without control subjects. Any study notmeasuring the diagnostic accuracy of the test usingan acceptable independent reference standard forcase definition.

APPENDIX 3: REVIEWERS

AAN Quality Standards Subcommittee Members.

Gary Franklin, MD, MPH—Co-Chair; CatherineZahn, MD—Co-Chair; Milton Alter, MD, PhD; Ste-


phen Ashwal, MD; Richard M. Dubinsky, MD; Jac-queline French, MD; Gary Friday, MD; MichaelGlantz, MD; Gary Gronseth, MD; Deborah Hirtz,MD; Robert G. Miller, MD; David Thurman, MD;and William Weiner, MD.

AANEM Practice Issue Review Panel Members. Rich-ard M. Dubinsky, MD, Chair, Michael T. Andary,MD, MS, Carmel Armon, MD, MHS, MS, William W.Campbell, MD, Joseph V. Campellone Jr., MD, Earl J.Craig, MD, Kenneth James Gaines, MD, JamesHoward Jr., MD, Robert G. Miller, MD, Atul Patel,MD, Yuen T. So, MD, PhD, and Robert A. Werner,MD, MS.

AAPM&R Guidelines Committee Members. HilarySiebens, MD, Chair, Greg Carter, MD, David Chen,MD, John Cianca, MD, Gerard Francisco, MD,Deanna Janora, MD, Bharat Patel, MD, GerardMalanga, MD, Jay Meythaler, MD, JD, Frank Salvi,MD, Richard Zorowitz, MD, and Maury Ellenberg,MD.

APPENDIX 4. DEFINITIONS FOR STRENGTH OFRECOMMENDATIONS

Level A: Established as effective, ineffective, or harm-ful for the given condition in the specified popula-tion. Usually, a Level A recommendation requiresthat the pooled result from two or more distinctClass I studies demonstrates a consistent, significant,and important effect.

Level B: Probably effective, ineffective, or harm-ful for the given condition in the specified popula-tion. Usually, a Level B recommendation requiresthat a single Class I study demonstrates a significantand important effect, or the pooled result from twoor more distinct Class II studies demonstrates a con-sistent, significant, and important effect.

Level C: Possibly effective, ineffective, or harmfulfor the given condition in the specified population.Usually, a Level C recommendation requires that asingle Class II study demonstrates a significant andimportant effect, or the pooled result of two or moredistinct Class III studies demonstrates a consistent,significant, and important effect.

Level U: Data that are inadequate or conflicting.Given the current knowledge the intervention is un-proven and an evidence-based recommendation can-not be made.

DISCLAIMER

The diagnosis of polyneuropathy is complex. Thecase definition is not intended to replace the clinical

judgment of experienced physicians in the diagnosisof polyneuropathy, because none of the criteria haveperfect diagnostic accuracy. This statement is pro-vided as an educational service of the AAN, theAAEM, and the AAPM&R. It is based on an assess-ment of current scientific and clinical information. Itis not intended to include all possible proper meth-ods of care for a particular neurological problem orall legitimate criteria for choosing to use a specificprocedure. Neither is it intended to exclude anyreasonable alternative methodologies. The AAN,AAEM, and AAPM&R recognize that specific caredecisions are the prerogative of the patient and phy-sician caring for the patient, based on all of thecircumstances involved.

REFERENCES

Strength of evidence is indicated for references usedto formulate case definition.1. American Association of Electrodiagnostic Medicine. Guide-

lines in electrodiagnostic medicine. Muscle Nerve 1999;22(Suppl 8):S3–S300.

2. Dyck PJ, Bushek W, Spring EM, Karnes JL, Litchy WJ, O’BrienPC, et al. Vibratory and cooling detection thresholds com-pared with other tests in diagnosing and staging diabeticneuropathy. Diabetes Care 1987;10:432–440. (Class II)

3. Dyck PJ, Davies JL, Litchy WJ, O’Brien PC. Longitudinalassessment of diabetic polyneuropathy using a compositescore in the Rochester Diabetic Neuropathy Study cohort.Neurology 1997;49:229–239. (Class III)

4. Dyck PJ, Karnes JL, Daube J, O’Brien P, Service FJ. Clinicaland neuropathological criteria for the diagnosis and stagingof diabetic polyneuropathy. Brain 1985;108:861–880.

5. Dyck PJ, Karnes JL, O’Brien PC, Litchy WJ, Low PA, MeltonLJ. The Rochester Diabetic Neuropathy Study: reassessmentof tests and criteria for diagnosis and staged severity. Neurol-ogy 1992;42:1164–1170.

6. Dyck PJ, Sherman WR, Hallcher LM, Service FJ, O’Brien PC,Grina LA, et al. Human diabetic endoneurial sorbital, fruc-tose and myo-inositol related to sural nerve morphometry.Ann Neurol 1980;8:590–596.

7. Feldman EL, Stevens MJ, Thomas PK, Brown MB, Canal N,Greene DA. A practical two-step quantitative clinical and elec-trophysiological assessment for the diagnosis and staging ofdiabetic neuropathy. Diabetes Care 1994;17:1281–1289.(Class III)

8. Franklin GM, Kahn LB, Baxter J, Marshall JA, Hamman RF.Sensory neuropathy in non-insulin-dependent diabetes melli-tus. The San Luis Valley Diabetes Study. Am J Epidemiol1990;131:633–643. (Class I)

9. Franse LV, Valk GD, Dekker JH, Heine RJ, van Eijk JTM.“Numbess of the feet” is a poor indicator for polyneuropathyin type 2 diabetic patients. Diabet Med 2000;17:105–110.(Class I)

10. Gentile S, Turco S, Corigliano, Marmo R, The SIMSDNGroup. Simplified diagnostic criteria for diabetic distal poly-neuropathy. Preliminary data of a multicentre study in theCampania region. Acta Diabetol 1995;32:7–12. (Class II)

11. Holland NR, Stocks A, Hauer P, Cornblath DR, Griffin JW,McArthur JC. Intraepidermal nerve fiber density in patientswith painful sensory neuropathy. Ann Neurol 1997;48:708–711.

12. Jones J, Hunter D. Consensus methods for medical and healthservices research. BMJ 1995;311:376–380.


13. Kennedy WR, Wendelschafer-Crabb G. The innervation ofhuman epidermis. J Neurol Sci 1993;115:184–190.

14. Lauria G, Sghirlanzoni A, Lombardi R, Pareyson D. Epider-mal nerve fiber density in sensory ganglionopathies: clinicaland neurophysiologic correlations. Muscle Nerve 2001;24:1034–1039.

15. Maser RE, Becker DJ, Drash AL, Ellis D, Kuller LH, GreeneDA, et al. Pittsburgh epidemiology of diabetes complicationsstudy. Measuring diabetic neuropathy follow-up results. Dia-betes Care 1992;15:525–527. (Class II)

16. McArthur JC, Stocks AE, Hauer P, Cornblath DR, Griffin JW.Epidermal nerve fiber density. Normative reference rangeand diagnostic efficiency. Arch Neurol 1998;55:1513–1520.

17. Meijer JWG, van Sonderen E, Blaauwwiekel EE, Smit AJ,Groothoff JW, Eisma WH, et al. Diabetic neuropathy exami-nation. A hierarchial scoring system to diagnose distal poly-neuropathy in diabetes. Diabetes Care 2000;23:750–753.(Class III)

18. Monticelli ML, Beghi E, the Italian General Practitioner StudyGroup (IGPSG). Chronic symmetric polyneuropathy in theelderly. A field screening investigation in two regions of Italy:background and methods of assessment. Neuroepidemiology1993;12:96–105. (Class II)

19. Peripheral Nerve Society. Diabetic polyneuropathy in con-trolled clinical trials: consensus report of the PeripheralNerve Society. Ann Neurol 1995;38:478–482.

20. Proceedings of a consensus development conference on stan-dardized measures in diabetic neuropathy. Neurology 1992;42:1823–1839.

21. Quantitative sensory testing: A consensus report from thePeripheral Neuropathy Association. Neurology 1993;43:1050–1052.

22. Rempel D, Evanoff B, Amadio PC, de Krom M, Franklin G,Franzblau A, et al. Consensus criteria for the classification ofcarpal tunnel syndrome in epidemiologic studies. Am J PublicHealth 1998;88:1447–1451.

23. Shy ME, Frohman EM, So YT, Arezzo JC, Cornblath DR,Giuliani MJ, et al. Therapeutics and technology assess-ment subcommittee of the American Academy of Neurol-ogy. Quantitative sensory testing. Report of the Therapeu-tics and Technology Assessment Subcommittee of theAmerican Academy of Neurology. Neurology 2003;60:898 –904.

24. Teunissen LL, Notermans NC, Franssen H, van der Graaf Y,Oey PL, Linssen WH, et al. Differences between hereditarymotor and sensory neuropathy type 2 and chronic idiopathicaxonal neuropathy. A clinical and electrophysiological study.Brain 1997;120:955–962. (Class III)

25. Valk GD, Nauta JJP, Strijers RLM, Bertelsmann FW. Clinicalexamination versus neurophysiological examination in thediagnosis of diabetic polyneuropathy. Diabetic Med 1992;9:716–721. (Class II)

26. Van Dijk GW, Wokke JHJ, Notermans NC, van Gijn J, Franssen H.Diagnostic value of myotatic reflexes in axonal and demyelinatingpolyneuropathy. Neurology 1999;53:1573–1576. (Class III)


AANEM PRACTICE PARAMETER ABSTRACT: Distal symmetric polyneuropathy (DSP) is the most commonvariety of neuropathy. Since the evaluation of this disorder is not standardized,the available literature was reviewed to provide evidence-based guidelinesregarding the role of autonomic testing, nerve biopsy, and skin biopsy for theassessment of polyneuropathy. A literature review using MEDLINE, EMBASE,Science Citation Index, and Current Contents was performed to identify the bestevidence regarding the evaluation of polyneuropathy published between 1980and March 2007. Articles were classified according to a four-tiered level ofevidence scheme and recommendations were based on the level of evidence.(1) Autonomic testing may be considered in the evaluation of patients withpolyneuropathy to document autonomic nervous system dysfunction (Level B).Such testing should be considered especially for the evaluation of suspectedautonomic neuropathy (Level B) and distal small fiber sensory polyneuropathy(SFSN) (Level C). A battery of validated tests is recommended to achieve thehighest diagnostic accuracy (Level B). (2) Nerve biopsy is generally accepted asuseful in the evaluation of certain neuropathies as in patients with suspectedamyloid neuropathy, mononeuropathy multiplex due to vasculitis, or with atyp-ical forms of chronic inflammatory demyelinating polyneuropathy (CIDP). How-ever, the literature is insufficient to provide a recommendation regarding whena nerve biopsy may be useful in the evaluation of DSP (Level U). (3) Skin biopsyis a validated technique for determining intraepidermal nerve fiber (IENF) den-sity and may be considered for the diagnosis of DSP, particularly SFSN (LevelC). There is a need for additional prospective studies to define more exactguidelines for the evaluation of polyneuropathy.

Muscle Nerve 39: 106–115, 2009

EVALUATION OF DISTAL SYMMETRIC POLYNEUROPATHY: THE

ROLE OF AUTONOMIC TESTING, NERVE BIOPSY, AND SKIN

BIOPSY (AN EVIDENCE-BASED REVIEW)

J .D . E N G L A N D , M D ,1 G .S . G R O N S E T H , M D ,2 G . F R A N K L IN , M D ,3 G .T . C A R T E R , M D ,4 L .J . K IN S E L L A , M D ,5 J .A . C O H E N , M D ,6

A .K . A S B U R Y , M D ,7 K . S Z IG E T I, M D , P H D ,8 J .R . L U P S K I, M D , P H D ,9 N . L A T O V , M D ,10 R .A . L E W IS , M D ,11 P .A . L O W , M D ,12

M .A . F IS H E R , M D ,13 D . H E R R M A N N , M D ,14 J .F . H O W A R D , M D ,15 G . L A U R IA , M D ,16 R .G . M IL L E R , M D ,17

M . P O L Y D E F K IS , M D ,18 a n d A .J . S U M N E R , M D 19 R e p o r t o f th e A m e r ic a n A c a d e m y o f N e u r o lo g y , th e A m e r ic a n A s s o c ia tio n

o f N e u r o m u s c u la r a n d E le c tr o d ia g n o s tic M e d ic in e , a n d th e A m e r ic a n A c a d e m y o f P h y s ic a l M e d ic in e a n d R e h a b ilita tio n

1 Louisiana State University Health Sciences Center, Baton Rouge, Louisiana, USA2 University of Kansas, Lawrence, Kansas, USA3 University of Washington, Seattle, Washington, USA4 Providence Health System, Southwest Washington, Seattle, Washington, USA5 Tenet-Forest Park Hospital, St. Louis, Missouri, USA6 Dartmouth Hitchcock Medical Center, Lebanon, New Hampshire, USA7 University of Pennsylvania Hospital, Philadelphia, Pennsylvania, USA8 Baylor College of Medicine, Houston, Texas, USA9 Baylor College of Medicine, Houston, Texas, USA10 Weill Medical College of Cornell, Ithaca, New York, USA11 Wayne State University School of Medicine, Detroit, Michigan, USA12 Mayo Clinic, Rochester, Minnesota

A b b r e v ia tio n s : AAN, American Academy of Neurology; AANEM, American Academy of Neuromuscular and Electrodiagnostic Medicine; AAPM&R, American Academyof Physical Medicine and Rehabilitation; CASS, composite autonomic scoring scale; CIDP, chronic inflammatory demyelinating polyneuropathy; CMT, Charcot–Marie–Tooth; DSP, distal symmetric polyneuropathy; EDX, electrodiagnostic; EFNS, European Federation of Neurological Societies; IENF, intraepidermal nerve fiber; NCS, nerveconduction study; QSART, quantitative sudomotor axon reflex test; SFSN, small fiber sensory polyneuropathy; TST, thermoregulatory sweat testingK e y w o r d s : prospective studies; evaluation; distal symmetric polyneuropathyC o r r e s p o n d e n c e to : American Association of Neuromuscular & Electrodiagnostic Medicine, 2621 Superior Drive NW, Rochester, MN 55901; e-mail:[email protected]

© 2008 Wiley Periodicals, Inc.Published online 15 December 2008 in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/mus.21227

10 6 AANEM Practice Parameter MUSCLE & NERV E J anuary 2009

13 Loyola University Chicago Stritch School of Medicine, Chicago, Illinois, USA14 University of Rochester Medical Center, Rochester, New York, USA15 University of North Carolina, Chapel Hill, North Carolina, USA16 National Neurological Institute “ Carlo Besta,” Milan, Italy17 California Pacifi c Medical Center, San Francisco, California, USA18 Johns Hopkins Medical Institute, Baltimore, Maryland, USA19 Louisiana State University Health Sciences Center, Baton Rouge, Louisiana, USA

Accepted 9 October 2008

INTRODUCTION

J u s tifi c a tio n . Polyneuropathy is a relatively com-mon neurological disorder.10 The overall prevalenceis !2,400 ( 2.4% ) per 100,000 population, but inindividuals older than 55 years the prevalence risesto !8,000 ( 8% ) per 100,000.9,32 Since there aremany etiologies of polyneuropathy, a logical clinicalapproach is needed for evaluation and management.

This practice parameter provides recommenda-tions for the evaluation of distal symmetric polyneu-ropathy ( DSP) based on a prescribed review andanalysis of the peer-reviewed literature. The param-eter was developed to provide physicians with evi-dence-based guidelines regarding the role of auto-nomic testing, nerve biopsy, and skin biopsy for theassessment of polyneuropathy. The diagnosis of DSPshould be based on a combination of clinical symp-toms, signs, and electrodiagnostic criteria as out-lined in the previous case defi nition.10 ( See MissionStatement, below, for details.)

Fo rm a tio n o f Ex p e r t Pa n e l. The PolyneuropathyTask Force included 19 physicians with representa-tives from the American Academy of Neurology( AAN) , the American Academy of Neuromuscularand E lectrodiagnostic Medicine ( AANE M) , and theAmerican Academy of Physical Medicine and Reha-bilitation ( AAPM& R) . All of the task force membershad extensive experience and expertise in the areaof polyneuropathy. Additionally, four members hadexpertise in evidence-based methodology and prac-tice parameter development. Three are currentmembers ( J.D.E ., G .S.G ., G .F.) , and one is a formermember ( R.G .M.) of the Q uality Standards Subcom-mittee ( Q SS) of the AAN. The task force developeda set of clinical q uestions relevant to the evaluationof DSP, and subcommittees were formed to addresseach of these q uestions.

DESCRIPTION OF THE ANALYTIC PROCESS

The literature search included O V ID ME DLINE( 1966 to March 2007) , O V ID E xcerpta Medica ( E M-BASE ; 1980 to March 2007) , and O V ID CurrentContents ( 2000 to March 2007) . The search in-

cluded articles on humans only and in all languages.The search terms selected were peripheral neurop-athy, polyneuropathy, and distal symmetric polyneu-ropathy. These terms were cross-referenced with theterms diagnosis, electrophysiology, autonomic test-ing, nerve biopsy, and skin biopsy.

Panel experts were asked to identify additionalarticles missed by the initial search strategy. Further,the bibliographies of the selected articles were re-viewed for potentially relevant articles.

Subgroups of committee members reviewed thetitles and abstracts of citations identifi ed from theoriginal searches and selected those that were poten-tially relevant to the evaluation of polyneuropathy.Articles deemed potentially relevant by any panelmember were also obtained.

E ach potentially relevant article was subseq uentlyreviewed in entirety by at least three panel members.E ach reviewer graded the risk of bias in each articleby using the diagnostic test classifi cation-of-evidencescheme ( Appendix 2) . In this scheme, articles attain-ing a grade of Class I are judged to have the lowestrisk of bias, and articles attaining a grade of Class IVare judged to have the highest risk of bias. Disagree-ments among reviewers regarding an article’s gradewere resolved through discussion. Final approval wasdetermined by the entire panel.

The Q uality Standards Subcommittee ( AAN) , thePractice Issues Review Panel ( AANE M) , and thePractice G uidelines Committee ( AAPM& R) ( Appen-dix 1A– C) reviewed and approved a draft of thearticle. The draft was next sent to members of theAAN, AANE M, and AAPM& R for further review andthen to Neurology for peer review. Boards of the AAN,AANE M, and AAPM& R reviewed and approved thefi nal version of the article. At each step of the reviewprocess, external reviewers’ suggestions were explic-itly considered. When appropriate, the expert panelmade changes to the document.

ANALYSIS OF THE EVIDENCE

The search yielded 1,045 references with abstracts.After reviewing titles and abstracts, 106 articles werereviewed and classifi ed.

AANEM Practice Parameter MUSCLE & NERV E J anuary 2009 10 7

Ro le o f Clin ic a l Au to n o m ic Te s tin g in th e Ev a lu a tio n o f

Po ly n e u r o p a th y . Autonomic nervous system dys-function occurs in several phenotypes. It may occuras one component of a generaliz ed polyneuropathysuch as DSP of diabetes. Such polyneuropathies areusually diagnosed by a combination of neuropathicsymptoms, decreased or absent ankle refl exes, de-creased distal sensation, distal muscle weakness oratrophy, and abnormal nerve conduction studies( NCSs) .10 The majority of these features constituteevidence of “ large fi ber” sensory and motor involve-ment. However, signs of autonomic nervous systeminvolvement may also constitute fi ndings indicativeof DSP. In DSP with autonomic involvement, themost common clinical fi ndings are abnormalities ofsweating and circulatory instability in the feet.9,10

A second phenotype is that of an autonomicneuropathy such as in amyloidosis and autoimmuneautonomic neuropathy, where autonomic nerves areaffected disproportionately relative to somaticnerves.29 In these neuropathies, autonomic fi berscan be affected in isolation and their involvementmay precede somatic fi ber involvement.47

A third relatively common phenotype is distalsmall fi ber sensory polyneuropathy ( SFSN) , whichcan manifest as burning pain affecting the feet, oftenwith allodynia and sometimes with erythromelalgia( red hot and painful skin) . Involvement of auto-nomic and somatic C fi bers usually occurs concur-rently in small fi ber polyneuropathy.47

Wh a t Is th e Us e fu ln e s s o f Clin ic a l Au to n o m ic Te s tin g in

th e Ev a lu a tio n o f Po ly n e u r o p a th y , a n d Wh ic h Te s ts

Ha v e th e Hig h e s t Se n s itiv ity a n d Sp e c ifi c ity ? Cur-rently available autonomic tests can provide indicesof cardiovagal, adrenergic, and postganglionic sudo-motor function. As such they provide indices forboth parasympathetic and sympathetic autonomicfunction. Heart rate variability testing is a simple andreliable test of cardiovagal function. It detects thepresence of diabetic polyneuropathy with nearly thesame sensitivity as NCSs ( Class II) .7 Specifi city is high( 97.5% ) for identifying parasympathetic defi cits ifthe recommended age-controlled values are used( Class II) .26 Intrinsic cardiac disease can affect theresults of this test, and this possibility must be con-sidered in the interpretation.

Cardiovagal function can be evaluated using dif-ferent indices in the time and freq uency domains.56

There is no compelling evidence that one method isbetter than another or that the use of multiple indi-ces confers any advantage. Heart rate variability todeep breathing is the most widely used test of car-

diovagal function and has a specifi city of !80%( Class II) .24

The vagal component of the barorefl ex can beevaluated by q uantitating the heart period responseto induced changes in blood pressure ( BP) . A well-studied test is the modifi ed O xford method.8 Thetest consists of an evaluation of heart period re-sponses to induced increases and decreases in arte-rial BP. The increase is evoked by intravenous phen-ylephrine and decrease by nitroprusside inincremental doses. Barorefl ex sensitivity is defi nedby the slope of the heart period to BP relationship.Linearity is req uired ( R " 0.85) . The advantage ofthis test is that it evaluates vagal barorefl ex sensitiv-ity; however, the disadvantage is that the test is inva-sive and not widely performed. Approximation ofthis method is possible by relating heart period al-terations to changes in BP induced by the V alsalvamaneuver.52 The sensitivity and specifi city of invasiveand noninvasive tests of barorefl ex function arehigh, but these tests are not generally used in thestudy of neuropathy since their value is consideredonly additive to current tests of cardiovagal function( Class II) .24,29,44,54

Thermoregulatory sweat testing ( TST) is a sensitivetest of sudomotor function that utiliz es an indicatorsubstance whose color changes upon exposure tosweat.12,30 The test results can be semiq uantitated byestimating the percentage of skin surface that is an-hidrotic. Since the test is tedious, messy, and time-consuming, it is not routinely done. Additionally, TSTis not able to distinguish between postganglionic,preganglionic, and central lesions.12,30 The most q uan-titative test of sudomotor function is the q uantitativesudomotor axon refl ex test ( Q SART) .25 Q SART is me-diated by impulses traveling antidromically then ortho-dromically along the postganglionic sympathetic sudo-motor axon. Q SART can detect distal sudomotor losswith a sensitivity of 75% – 90% ( Class III) .26,35,50,51 Sev-eral studies have demonstrated that Q SART can deter-mine sudomotor abnormalities with relatively highsensitivity and specifi city in many types of polyneurop-athies ( Class II and III) .7,25,26,28,29,31,35,47,50 In threeClass III studies, Q SART was shown capable of detect-ing distal small fi ber polyneuropathy with a sensitivityof "75% .35,50,51

Skin vasomotor refl exes assessed by monitoringskin blood fl ow using laser Doppler fl owmeter hasnot been well studied. Limited data from one ClassIII study using this techniq ue demonstrated an un-acceptably large coeffi cient of variation.27

Analysis of the available Class II and III studies onautonomic testing indicate that a combination ofautonomic refl ex screening tests provides distinct

10 8 AANEM Practice Parameter MUSCLE & NERV E J anuary 2009

advantages over single modality methods. The com-posite autonomic scoring scale ( CASS) , which in-cludes Q SART, orthostatic blood pressure, heart rateresponse to tilt, heart rate response to deep breath-ing, the V alsalva ratio, and beat-to-beat BP measure-ments during Phases II and IV of the V alsalva ma-neuver, tilt, and deep breathing provides a useful10-point scale of autonomic function ( Class II) .24,29

In a study of 78 patients with graded autonomicfailure obtained by selecting approximately eq ualnumbers of patients with multiple system atrophy,Parkinson’s disease, autonomic neuropathies, andidiopathic peripheral neuropathies, this combina-tion of tests provided a noninvasive, sensitive, spe-cifi c, and reproducible methodology for grading thedegree of autonomic dysfunction ( Class II) .24

Conclusions. Autonomic testing is probably use-ful in documenting autonomic nervous system in-volvement in polyneuropathy ( Class II and III) . Thesensitivity and specifi city vary with the particular test.The utiliz ation of the combination of autonomicrefl ex screening tests in the CASS provides the high-est sensitivity and specifi city for documenting auto-nomic dysfunction ( Class II) .

Recommendations. Autonomic testing should beconsidered in the evaluation of patients with poly-neuropathy to document autonomic nervous systeminvolvement ( Level B) . Autonomic testing should beconsidered in the evaluation of patients with sus-pected autonomic neuropathies ( Level B) and maybe considered in the evaluation of patients with sus-pected distal SFSN ( Level C) . The combination ofautonomic screening tests in the CASS should beconsidered to achieve the highest diagnostic accu-racy ( Level B) . If the full battery of tests in the CASSis not available, a combination of tests of cardiovagalfunction ( e.g., heart rate response to deep breath-ing) and some test of adrenergic function may beconsidered as an alternative ( Level C) .24

Ro le o f Ne r v e Bio p s y in th e Ev a lu a tio n o f Po ly n e u r o p -

a th y . Nerve biopsy is generally accepted as useful inthe diagnosis of infl ammatory diseases of nervessuch as vasculitis, sarcoidosis, CIDP, infectious dis-eases such as leprosy, or infi ltrative disorders such astumor or amyloidosis.9 Nerve biopsy is most valuablein mononeuropathy multiplex or suspected vascu-litic neuropathy. There are no studies regarding therole of nerve biopsy in the evaluation of DSP, al-though on occasion the above-noted diseases maypresent in that fashion.

Wh a t Is th e Us e fu ln e s s o f Ne r v e Bio p s y in De te r m in in g

th e Etio lo g y o f Dis ta l Sy m m e tr ic Po ly n e u r o p a th y ?

O ut of 50 articles judged to be relevant, no articleattained a grade greater than Class IV . Most of thearticles discussed the nerve biopsy fi ndings in spe-cifi c diseases, the clinical suspicion of which hadprompted the biopsy.1– 3,5,6,13,14,34,39,40 – 42 No articleprovided guidance regarding when to perform anerve biopsy in the evaluation of DSP.

Conclusions. There is no evidence to support orrefute a conclusion regarding the role of nerve bi-opsy in the evaluation of DSP ( Class IV ) .

Recommendations. No recommendations can bemade regarding the role of nerve biopsy in deter-mining the etiology of DSP ( Level U) .

Ro le o f Sk in Bio p s y in th e Ev a lu a tio n o f Po ly n e u r o p a th y .

Skin biopsy is being increasingly used to evaluatepatients with polyneuropathy. The most commontechniq ue involves a 3-mm punch biopsy of skinfrom the leg. After sectioning by microtome, thetissue is immunostained with anti-protein-gene-prod-uct 9.5 ( PG P 9.5) antibodies and examined withimmunohistochemical or immunofl uorescent meth-ods. This staining allows for the identifi cation andcounting of intraepidermal nerve fi bers ( IE NF) .PG P 9.5 immunohistochemistry has been validatedas a reliable method for IE NF density determinationwith good intra- and interobserver reliability in nor-mal controls and patients with DSP.15,20,33,48

In March 2005 the E uropean Federation of Neu-rological Societies ( E FNS) published a guideline onthe use of skin biopsy in peripheral neuropathy.20

This comprehensive review focused on the technicalaspects of skin biopsy as well as normative data andcorrelations with other clinical, physiologic, andpathologic tools. The E FNS concluded that skin bi-opsy is a safe, validated, and reliable techniq ue forthe determination of IE NF density. The major con-clusion was that skin biopsy ( IE NF density) was di-agnostically effi cient at distinguishing polyneurop-athy patients ( including small fi ber neuropathy)from normal controls. The E FNS guideline also re-viewed the literature on IE NF morphologic changessuch as axonal swellings as a measure of distal sym-metric polyneuropathy.16,20,21 The E FNS concludedthat axonal swellings may be predictive of progres-sion of polyneuropathy but further studies wereneeded to determine their diagnostic accuracy.20

Wh a t Is th e Us e fu ln e s s a n d Dia g n o s tic Ac c u r a c y o f Sk in

Bio p s y in th e Ev a lu a tio n o f Po ly n e u r o p a th y ? Beyonddistinguishing asymptomatic normals from polyneu-ropathy patients, one clinical q uestion not addressedby the E FNS guideline was the diagnostic accuracy ofskin biopsy in distinguishing symptomatic patients

AANEM Practice Parameter MUSCLE & NERV E J anuary 2009 10 9

T a b le 1. Evidence table for autonomic testing.

Reference Y ear Target disorder Predictor

Reference

standard Cases Controls Design Spectrum Mask ed Class Sens Spec

50 1992 Distal small fiber

neuropathy

QSART,

QST,

H RV

Neurologic

exam and

EDx

4 0 129 Retrospective

Review

N N III 80% 72%

7 1992 Diabetic PN QAE EDx 3 80 3 57* Concurrent

comparative

B U nmask ed/

Independent

II QAE:97% "90%

26 1997 PN, Park inson’s,

multisystem

atrophy

QSART Older scale 18 557 Concurrent

comparison

B U nmask ed/

Independent

II "90% "90%

51 1999 Peripheral (small

fiber) neuropathy

QSART,

QST,

Clinical

Symptoms

EDx 13 8 3 57* (per Dr.

L ow)

Concurrent

comparative

B U nmask ed/

Independent

III QSART:80% ;

QST:67%

93 %

3 5 2001 Painful neuropathy QSART,

ART,

CASS

Clinical

evaluation

126 3 57* (per Dr.

L ow)

Non-comparative N N III ART:93 % ;

QSART:

73 %

94 %

24 1993 Diabetic PN CASS EDx and

standard

clinical

exam

78 3 50 Concurrent

comparative

N U nmask ed/

Independent

II "90% "90%

4 4 2007 Adrenergic

autonomic failure

B RSI MSNA 84 29 Concurrent

comparative

B U nmask ed/

Independent

II 86% "90%

54 2005 Multi system

atrophy,

peripheral

neuropathy

PRT,

CASS

Clinical exam 162 3 2 Concurrent

comparative

B U nmask ed/

Independent

II "90% "90%

4 7 2004 DSFN, PN, DN, IAN CASS Neurological

exam

11 3 8 Concurrent

comparative

N U nmask ed/

Independent

III 95% 90%

Ac ro n y m s D is ta l s m a ll fi b e r n e u ro p a th y : D S F N; D ia b e tic p e rip h e ra l n e u ro p a th y : D PN; Pe rip h e ra l n e u ro p a th y : PN; Q u a n tita tiv e s u d o m o to r a x o n re fl e x te s tin g : Q S ART; Co m p o s ite a u to n o m ic s e v e rity s c o re : CAS S ;

Q u a n tita tiv e a u to n o m ic e x a m in a tio n : Q AE; Id io p a th ic a u to n o m ic n e u ro p a th y : IAN; Q u a n tita tiv e s e n s o ry te s tin g : Q S T; Au to n o m ic re fl e x te s tin g : ART; Ele c tro d ia g n o s is : Ed x ; H e a rt ra te v a ria b ility : H RV ; Q u a n tita tiv e

s w e a t te s tin g ; B a ro re fl e x s e n s itiv ity in d e x : B RS I; Mu s c le s y m p a th e tic n e rv e a c tiv ity : MS NA; B lo o d p re s s u re re c o v e ry tim e : PRT.

110

AA

NE

MP

ractic

eP

ara

mete

rM

US

CLE

&N

ER

VE

January

2009

with polyneuropathy from symptomatic patientswithout polyneuropathy. For example, in patientswith painful feet, would skin biopsy accurately dis-tinguish patients with polyneuropathy from patientswith other conditions causing painful feet?

To address this separate q uestion, a subgroup ofthe Polyneuropathy Task Force ( J.D.E ., R.A.L., D.H.,G .L., M.P., and G .S.G .) independently reviewed theliterature regarding the diagnostic accuracy of skinbiopsy in DSP and in the SFSN form of DSP. To beconsidered for review, studies needed to determineIE NF density in patients with and without polyneu-ropathy. Furthermore, the data from studies had tobe presented in such a way as to allow calculation ofthe sensitivity and specifi city of skin biopsy for poly-neuropathy.

Nine studies met inclusion criteria.4,16 – 18,21–

22,33,36 – 37 O ne was a prospective cohort survey ofpatients presenting with bilateral painful feet andnormal strength, but skin biopsy was done only inthose with normal NCS.36 Patients with reducedIE NF density and normal NCS were assumed to havepainful small fi ber neuropathies. However, the studydid not compare the results of the IE NF density to anindependent reference standard to confi rm the pres-ence of small fi ber neuropathy. Thus, for the pur-poses of determining the diagnostic accuracy of skinbiopsy for polyneuropathy, this study was gradedClass IV .

The remaining studies employed a case-controldesign.16,18,21,22,33 In these studies the investigatorsperformed skin biopsies on patients with establishedpolyneuropathy and normal controls. No study in-cluded patients with conditions causing lower ex-tremity pain or sensory complaints that might beconfused with polyneuropathy. Thus, all studies hadpotential spectrum bias. Following the evidence clas-sifi cation scheme for studies of diagnostic accuracy,all of these studies were graded Class III.

All of the case-control studies showed a signifi -cant reduction in IE NF density in polyneuropathypatients as compared to controls.16,18,21,22,33 The sen-sitivity of decreased IE NF density for the diagnosis ofpolyneuropathy was moderate to good ( range 45% –90% ) . The specifi city of normal IE NF density for theabsence of polyneuropathy was very good ( range95% – 97% ) . Thus, the absence of reduced IE NF den-sity ( using the clinical impression as the diagnosticreference standard) would not “ rule out” polyneu-ropathy, but the presence of reduced IE FN densitywould importantly raise the likelihood of polyneu-ropathy.

The form of DSP for which IE NF assessment isparticularly diagnostically attractive is SFSN for the fol-

lowing reasons: ( 1) IE NF are the nerve terminals ofsomatic unmyelinated C fi bers, which are hypothesiz edto be predominantly affected in SFSN; ( 2) There hasbeen a lack of a direct objective measure of small fi bersensory nerves since objective measures of large fi berfunction ( e.g., NCS) are by most defi nitions normal inSFSN19; 3) Patients in whom SFSN is clinically sus-pected manifest with symptoms of small fi ber sensorydysfunction ( e.g., tingling, numbness, and neuropathicpain) but few objective signs, making it diffi cult todiagnose and to distinguish SFSN from nonneurologi-cal causes of sensory complaints.19

Since no validated objective gold standard existsfor the diagnosis of SFSN, the authors consideredwhether demonstration of a pathologic lesion ( smallsensory fi ber pathology on skin biopsy) should bethe de facto diagnostic standard or whether a clinicalimpression of SFSN should be the independent ref-erence standard. For the purposes of this parameter,a clinical impression of SFSN was adopted as theindependent reference standard for calculation ofsensitivity and specifi city of IE NF density in the de-tection of SFSN.

In order to assess the diagnostic accuracy of IE NFdensity assessment for SFSN, the literature was sur-veyed for studies assessing IE NF density in subjectswith clinically suspected SFSN ( symptoms or symp-toms and signs of DSP but with normal NCS) andcontrols where the diagnostic accuracy of IE NF den-sity for clinically defi ned SFSN could be determined.Four Class III studies met these criteria.18,23,46,55 Thesensitivity of IE NF density assessment at the ankle forDSP with normal NCS was 58% ( 20% for subjectswith symptoms but no signs of SFSN; 100% for sub-jects with symptoms and signs of SFSN) ,46 90% ,18

and 24% .23 In these studies, the specifi city of the testranged from 95% – 97.5% .18,23,46 The other case-con-trol study found that among patients with symptomsof SFSN and an abnormal pinprick examination inthe feet, but normal ankle refl exes, normal vibrationsensibility, and normal NCS that an IE NF density of#8 fi bers/ mm at the dorsal foot provided a sensitiv-ity of 88% , a specifi city of 91% , a positive predictivevalue of 0.9, and a negative predictive value of 0.83for the diagnosis of SFSN.55

Conclusions. IE NF density assessment using PG P9.5 immunohistochemistry is a validated, reproduc-ible marker of small fi ber sensory pathology. Skinbiopsy with IE NF density assessment is possibly use-ful to identify DSP that includes SFSN in symptom-atic patients with suspected polyneuropathy. ( ClassIII) .

Recommendations. For symptomatic patientswith suspected polyneuropathy, skin biopsy may be

AANEM Practice Parameter MUSCLE & NERV E J anuary 2009 111

considered to diagnose the presence of a polyneu-ropathy, particularly SFSN ( Level C) .

RECOMMENDATIONS FOR FUTURE RESEARCH

This comprehensive review reveals several weak-nesses in the current approach to the evaluation ofpolyneuropathy and highlights opportunities for re-search.

● Auton om ic testin g. Autonomic testing can, with ahigh degree of accuracy, document autonomicsystem dysfunction in polyneuropathy. This isparticularly relevant to small fi ber polyneurop-athy and the autonomic neuropathies. Researchis necessary to determine whether the docu-mentation of autonomic abnormalities is impor-tant in modifying the evaluation and treatmentof polyneuropathy. Specifi c tests such as Q SARTcan document small fi ber ( i.e., sudomotoraxon) loss with a high degree of sensitivity, mak-ing the test useful to confi rm the diagnosis ofsmall fi ber polyneuropathy. Since skin biopsywith determination of IE NF density can alsodocument small fi ber loss, there is a need forstudies that compare and correlate the two tech-niq ues.

● Nerv e biopsy. There are no studies of nerve bi-opsy in the evaluation of DSP. Although itwould be useful to know the outcome of well-designed prospective studies in this area, it isunlikely that such studies will be done.

● S k in biopsy. Skin biopsy with determination ofIE NF density is a techniq ue that has come ofage for the objective documentation of smallfi ber loss. This techniq ue provides a uniq ueopportunity for research in different varieties ofneuropathy. Further studies are needed to char-acteriz e the diagnostic accuracy of skin biopsyin distinguishing patients with suspected poly-neuropathy, particularly SFSN, from patientswith sensory complaints or pain unrelated toperipheral neuropathy. Prospective studies withappropriate “ other disease” controls should bedone to assess the sensitivity, specifi city, andpredictive values of IE NF density measurementto identify SFSN in patients with lower extremitypain or sensory complaints. A predeterminedindependent reference standard for the diagno-sis of SFSN should be specifi cally stated in suchstudies.

● A case defi nition of SFSN should be developed.Investigators need to determine whether thiscase defi nition should be based on clinical cri-

teria, pathological criteria ( e.g., skin biopsy) , ora combination of clinical, paraclinical, andpathologic criteria.

● The diagnostic accuracy of morphologicchanges ( e.g., axonal swellings) in the diagnosisof SFSN versus healthy controls and disease con-trols needs to be better defi ned.

● Studies exploring other uses for skin biopsybeyond identifi cation and q uantifi cation of DSPand SFSN have been reported and should befurther explored. Biopsies of glabrous skin anddermal skin include myelinated nerve fi bers,and have been shown to have potential utility inthe diagnosis of immune-mediated neuropa-thies, Charcot– Marie– Tooth ( CMT) , and re-lated diseases.22 O ther studies have employedskin biopsy for detection or monitoring of lep-rosy, hereditary amyloidosis, vasculitic neuropa-thy, and Fabry’s disease.49,53 Additional studiesare req uired to determine the usefulness of skinbiopsy in the diagnosis and monitoring of theseand other varieties of neuropathy.

● Serial IE NF density measurements and IE NFregenerative capacity are being studied andused as outcome measures in therapeutic tri-als.38,43 Further studies are needed to validateand determine the value of skin biopsy for thispurpose.

Mis s io n Sta te m e n t. The AAN, the AANE M, and theAAPM& R determined that there was a need for anevidence-based and clinically relevant practice pa-rameter for the evaluation of polyneuropathy. As aprelude to this project, the three organiz ations de-veloped a formal case defi nition of distal symmetricpolyneuropathy DSP.10 As outlined in this previouspublication, the most accurate diagnosis of distalsymmetric polyneuropathy is provided by a combi-nation of neuropathic symptoms, signs, and electro-diagnostic ( E DX ) studies. Since E DX studies aresensitive, specifi c, and validated measures of thepresence of polyneuropathy and can distinguish be-tween demyelinating and axonal types of neuropa-thy, they should be included as an integral part ofthe diagnosis.10 This practice parameter assumesthat a clinical diagnosis of polyneuropathy has beendetermined based on such criteria.

D isclaimer . The diagnosis and evaluation ofpolyneuropathy is complex. The practice parameteris not intended to replace the clinical judgment ofexperienced physicians in the evaluation of polyneu-ropathy. The particular kinds of tests utiliz ed by aphysician in the evaluation of polyneuropathy de-

112 AANEM Practice Parameter MUSCLE & NERV E J anuary 2009

pend on the specifi c clinical situation and the in-formed medical judgment of the treating physician.

This statement is provided as an educational ser-vice of the AAN, AANE M, and the AAPM& R. It isbased on an assessment of current scientifi c andclinical information. It is not intended to include allpossible proper methods of care for a particularneurologic problem or all legitimate criteria forchoosing to use a specifi c test or procedure. Neitheris it intended to exclude any reasonable alternativemethodologies. The AAN, AANE M, and AAPM& Rrecogniz e that specifi c care decisions are the prerog-ative of the patient and physician caring for thepatient, based on all of the circumstances involved.

Co n fl ic t o f In te r e s t. The AAN, AANE M, andAAPM& R are committed to producing independent,critical, and truthful clinical practice guidelines( CPG s) . Signifi cant efforts are made to minimiz e thepotential for confl icts of interest to infl uence therecommendations of this CPG . To the extent possi-ble, the AAN, AANE M, and AAPM& R keep separatethose who have a fi nancial stake in the success orfailure of the products appraised in the CPG s andthe developers of the guidelines. Confl ict of interestforms were obtained from all authors and reviewedby an oversight committee prior to project initiation.AAN, AANE M, and AAPM& R limit the participationof authors with substantial confl icts of interest. TheAAN, AANE M, AAPM& R forbid commercial partici-pation in, or funding of, guideline projects. Drafts ofthe guideline have been reviewed by at least threeAAN committees, AANE M and AAPM& R commit-tees, a network of neurologists, Neurology peer review-ers, and representatives from related fi elds. TheAAN G uideline Author Confl ict of Interest Policycan be viewed at www.aan.com.

APPENDIX 1 A

Q u a lity Sta n d a r d s Su b c o m m itte e (AAN). Jacq uelineFrench, MD, FAAN ( co-chair) ; G ary S. G ronseth, MD( co-chair) ; Charles E . Argoff, MD; E ric Ashman, MD;Stephen Ashwal, MD, FAAN ( ex-offi cio) ; Christo-pher Bever Jr., MD, MBA, FAAN; John D. E ngland,MD, FAAN ( Q SS facilitator) ; G ary M. Franklin, MD,MPH, FAAN ( ex-offi cio) ; Deborah Hirtz , MD ( ex-offi cio) ; Robert G . Holloway, MD, MPH, FAAN;Donald J. Iverson, MD, FAAN; Steven R. Messé, MD;Leslie A. Morrison, MD; Pushpa Narayanaswami,MD, MBBS; James C. Stevens, MD, FAAN ( E x-O ffi -cio) David J. Thurman, MD, MPH ( ex-offi cio) ; Sam-uel Wiebe, MD; Dean M. Wingerchuk, MD, MSc,FRCP( C) ; and Theresa A. Z esiewicz , MD, FAAN.

APPENDIX 1 B

Pra c tic e Is s u e s Re v ie w Pa n e l (AANEM). Yuen T. So,MD, PhD ( chair) ; Michael T. Andary, MD; Atul Pa-tel, MD; Carmel Armon, MD; David del Toro, MD;E arl J. Craig, MD; James F. Howard, MD; Joseph V .Campellone Jr., MD; Kenneth James G aines, MD;Robert Werner, MD; Richard Dubinsky, MD.

APPENDIX 1 C

Pra c tic e Gu id e lin e s Co m m itte e (AAPM& R). DexanneB. Clohan, MD ( chair) ; William L. Bockenek, MD;Lynn G erber, MD; E dwin Hanada, MD; Ariz R.Mehta, MD; Frank J. Salvi, MD, MS; and Richard D.Z orowitz , MD.

APPENDIX 2

Cla s s ifi c a tio n o f Ev id e n c e fo r Stu d ie s o f Dia g n o s tic Ac -

c u r a c y . Class I. E vide

distal symmetric polyneuropathy · 2011. 2. 24. · oduct: pp08 ‐ urse descript istal symmetri...

Documents