Residual fatigue in Guillain-Barre syndrome is related to axonal loss

Judith Drenthen, MD Bart C. Jacobs, MD, PhD Ellen M. Maathuis, MD Pieter A. van Doorn, MD, PhD Gerhard H. Visser, MD, PhD Joleen H. Blok, PhD

ABSTRACT

Correspondence to Dr. Drenthen !.drenthen"eras#us#c.nl

%b!ective To determine the occurrence of residual loss of peripheral nerve axons by motor unit number estimation (MUNE) and conventional nerve conduction studies (NCS) in patients with and without severe fati ue! Methods Thirty"nine patients at a median of # years (ran e $"%& years) after dia nosis of 'uillain"(arre syndrome were neurolo ically examined and divided in % sub roups based on the presence of severe fati ue (defined as a fati ue severity score )*)! +ll patients were investi ated with standard NCS and MUNE! Normal values for MUNE were collected in $, healthy controls! -esults. MUNE of the thenar muscles was lower in the $* patients with severe fati ue (median $%*/ inter0uartile ran e 1*"$,$) compared with the %, patients without severe fati ue (median %*#/ inter0uartile ran e $%2"&,*) (p 3 2!22%)! 4n the healthy controls/ MUNE was &*# (%,*" ,$1)! Severe fati ue was also related to lower sensory nerve action potential amplitude of the median (p 3 2!2$) and ulnar nerve (p 3 2!2&)! The % sub roups did not differ re ardin neurolo ic deficits/ disability/ and the remainin conventional motor NCS! Conclusion This study demonstrates that severe fati ue after 'uillain"(arre syndrome is related to more pronounced axonal loss/ represented by lower MUNEs and lower sensory nerve action potentials! Neurology5 2013&81 1827-1831
GLOSSARY CMAP 3 compound muscle action potential6 FSS 3 7ati ue Severity Scale6 GBS 3 'uillain"(arr1 syndrome6 HADS 3 8ospital +nxiety and 9epression Scale6 IQR 3 inter0uartile ran e6 MRC 3 Medical -esearch Council6 M 3 motor unit6 M !" 3 motor unit number estimation6 M P 3 motor unit potential6 !CS 3 nerve conduction studies6 #DSS 3 overall disability sum score6 S!AP 3 sensory nerve action potential!

Supplemental data at www.neurology.org

Guillain'Barre s(ndro#e )GB*+ is a subacute peripheral neuropath( that #a( cause severe ,eak ness.-./ Despite recover( o$ this ,eakness in the #a!orit( o$ patients, appro0i#atel( 123 have severe $ati4ue. 5his $ati4ue is #ore severe than in health( persons and #a( result in substantial disabilit(, ,ith a hi4h i#pact on perceived health status.6 5he #echanis#s underl(in4 post'GB* $ati4ue are unclear. 7 Previous research has sho,n that severe $ati4ue a$ter GB* is not related to recover( o$ #uscle stren4th and sensor( de$icits, nor to the t(pe o$ antecedent in$ection. 8 As a result, $ati4ue is so#eti#es attributed to stress or ps(cholo4ical $actors. 9n other neuro#uscular disorders, #ore severel( a$$ected patients e0perience a hi4her level o$ $ati4ue than less a$$ected patients.: 5his #a( su44est a relation bet,een e0perienced $ati4ue and a0onal loss. Ho,ever, nerve conduction studies );C*+ in patients ,ith GB* have not sho,n a relation bet,een the occurrence o$ $ati4ue and a0onal loss. < A $ollo,'up stud( in patients ,ho had GB* de#onstrated that residual a0onal loss is present in a nu#ber o$ clinicall( ,ell'recovered patients. 1 Ho,ever, ,hether this a0onal loss is associated ,ith $ati4ue is unkno,n. Advanced electroph(siolo4ic techni=ues such as #otor unit nu#ber esti#ation )M>;E+ #i4ht be #ore sensitive in detectin4 a0onal loss than conventional electroph(siolo4ic #ethods, because ,ith conventional #ethods, a0onal loss is #asked b( reinnervation processes. 1 Hence, the ai# o$ the present stud( ,as to use M>;E, a techni=ue that provides in$or#ation on the
___________ ___________________ _________________________________________________________________ ______
?ro# the Depart#ents o$ Clinical ;europh(siolo4( )J.D., G.H.V., J.H.B.+, ;eurolo4( )J.D., B.C.J., PA.v.D.+, and 9##unolo4( )B.C.J.+, Eras#us MC, >niversit( Mcdical Center @otterda#, the ;etherlands. Go to ;eurolo4(.or4 $or $ull disclosures. ?undin4 in$or#ation and disclosures dee#ed relevant b( the authors, i$ an(, are provided at the end o$ the article.

: %2$& +merican +cademy of Neurolo y

1BZ 7

nu#ber o$ $unctional #otor units )M>s+, to investi4ate the relation bet,een a0onal loss and severe $ati4ue a$ter GB*.
M"$H#DS Su%&e'ts( 5hirt( 'nine GB* patients )/2 #en, -A
,o#en, a4e ran4e 6-'<< (ears+ ,ere included in this stud(. Patients ,ere recruited $ro# an e0istin4 database o$ GB* patients included in previous clinical studies.A'-- ?ro# this database, -/A patients ,ere approached to participate in the current stud( and 72 )663+ a4reed. 5,o patients ,ere re$erred to us b( neurolo4ists $ro# re4ional hospitals $or the purpose o$ this stud(. 5hree o$ the 7/ patients that ,ere ,illin4 to participate in this stud( had conco#itant diseases )pul#onar( proble#s, diabetes, cancer+ and ,ere not included in this stud(. Ail patients $ul$illed the dia4nostic criteria $or GB* -/ at the ti#e o$ onset o$ disease, and ,ere clinicall( stable $or at least - (ear. 5he GB* ,as not acco#panied b( conco#itant conditions either in the acute phase or at $ollo,'up )e.4., no diabetes, #ali4nanc(, h(poth(roidis#, cardiac and pul#onar( proble#s, chronic diseases, or depression+, and the patients did not use #edication that #i4ht cause or in$luence $ati4ue. 5he #edian ti#e bet,een the onset o$ GB* and enroll#ent in this stud( ,as 1 (ears )ran4e -'/6 (ears+. ?ourteen health( sub!ects ,ith a si#ilar distribution o$ se0 and a4e )< #en, < ,o#en, a4e /<'<6 (ears+ ,ere recruited as controls. Be$ore enterin4 the stud(, all participants ,ere clinicall( and electroph(sioio4icall( screened $or carpal tunnel s(ndro#e. 5he patients under,ent a $ull neurolo4ic investi4ation )usin4 clinical scores as described belo,+, standard ;C*, and M>;E #easure#ents. 9n the health( sub!ects, onl( M>;E #easure#ents ,ere per$or#ed.

sa#ple increases the accurac( o$ die esti#ate. //,/6 9n this stud(, ,e deter#ined the M>;E o$ die thenar #uscles ,ith the adapted #ultiple'point sti#ulation techni=ue usin4 hi4h'densit( sur$ace EMG./6 9n hi4h'densit( sur$ace EMG, the M> responses are recorded ,ith an arra( o$ -/: densel( spaced electrodes. >sin4 such an arra( provides spatiote#poral pro$iles )D$in4erprintsE+ o$ individ' ual M>s. Because the $in4erprint o$ each M> is uni=ue, it $acilitates the detection o$ M>Ps $ro# sin4le M>s. 9n turn, this increases the nu#ber o$ M>s that can be sa#pled co#pared ,ith conventional sin4le'electrode recordin4s. 5he recordin4s co##enced ,ith sti#ulation o$ the #edian nerve at an intensit( lo, enou4h to elicit a sin4le, all'or'none M> response. 5hen, the sti#ulus intensit( ,as 4raduall( increased until another response, representin4 the coactivation o$ a second M>, ,as seen and then slo,l( to still hi4her intensities, activatin4 ever #ore M>s, until the variation in the si4nal ,as too hi4h to enable reliable o$$line anal(sis. 5he sti#ulus electrode ,as then #oved to a sli4htl( di$$erent location alon4 the nerve and the process ,as repeated. >suall(, this allo,ed the recordin4 o$ responses $ro# other M>s. 5he entire process ,as repeated until appro0i#atel( /2 to 62 M>Ps ,ere sa#pled. ?ro# this sa#ple, the #ean M>P ,as calculated, ,hich ,as divided into the #a0i#u# CMAP to derive the M>;E./6./.F

Statisti's( All data ,ere $irst tested $or nor#alit( usin4 the
Gol#o4orov'*#irnov test. ;or#all( distributed data )a4e and (ears since GB* dia4nosis+ ,ere presented as #eans and *Ds and nonnor#all( distributed and ordinal data )M>;E, clinical scales, and ;C* para#eters+ as #edians and inter=uartile ran4es )9H@s+. ;or#all( distributed data ,ere anal(Ced usin4 the /' sided t test $or / independent sa#ples or the -',a( anal(sis o$ variance $or 6 independent sa#ples. ;onnor#all( distributed data and ordinal data ,ere anal(Ced ,ith the Mann'Ihitne( U test. 5he J/ test ,Ks used $or proportions. Correlations ,ere tested ,ith the *pear#an test. *tep,ise #ultiple re4ression anal(sis ,as per$or#ed to investi4ate con$oundin4 $actors on the ?**. All calculations ,ere per$or#ed usin4 *P** -8.2 )*P** 9nc., Chica4o, 9L+. A p value M2.28 ,as considered to be statisticall( si4ni$icant.

Clini'al s'ores( 5o deter#ine the residual e$$ects o$ GB*, ,e re'

corded the GB* disabilit( score,
M -6

overall disabilit( su# score

-8

)%D**+, and Medical @esearch Council )M@C+ su# score.

5he

presence and severit( o$ $ati4ue ,ere assessed ,ith the ?ati4ue *everit( *cale )?**+.-: 5he ?** is a validated sel$'reported $ati4ue =uestionnaire containin4 A ite#s, each o$ ,hich is scored on a scale o$ - to <. *evere $ati4ue ,as de$ined accordin4 to previousl( established criteria as a #ean ?** score o$ 8.2 or #ore. 8 5o e0clude the potential e$$ect o$ )subclinical+ depression on e0perienced $ati4ue, patients ,ere additionall( screened ,ith the Hospital An0iet( and Depression *cale )HAD*+.-< 5he HAD* contains -7 ite#s and consists o$ / subscales an0iet( and depression. Each ite# is rated on a 7'point scale, 4ivin4 #a0i#u# scores o$ /-. %n the depression subscale, a score o$ 2 to < #a( be considered nor#al.
-<

Standard )roto'ol a))ro*als+ registrations+ and )atient 'onsents( 5he local Medical Ethics Co##ittee approved the
e0peri#ental protocol. All sub!ects 4ave ,ritten in$or#ed consent.

R"S ,$S 5he 6A GB* patients ,ere divided into /

Standard !CS( *tandard #otor ;C* ,ere per$or#ed o$ the ulnar,

#edian, peroneal, and ribial nerves. *ensor( ;C* ,ere per$or#ed o$ the ulnar, #edian, and sural nerves. 9$ distal li#b te#peratures ,ere belo, 6/BC, the ar#s and le4s ,ere ,ar#ed $or 62 #inutes. ls *ti#ulation and recordin4s sites ,ere standardiCed.
,

?or #otor

nerves, the distal and pro0i#al baseline'peak co#pound #uscle action potential )CMAP+ a#plitudes, distal #otor latenc(, #otor nerve conduction velocit(, and ?',ave latencies ,ere deter#ined. ?or sensor( nerves, the baseline'peak sensor( nerve action potential )*;AP+ a#plitude and sensor( nerve conduction velocit( ,ere #easured. @e$erence values ,ere derived $ro# Buschbacher and Prahlo,.-A

Motor unit num%er estimation( M>;E is based on the division o$

the #a0i#al CMAP a#plitude b( an esti#ate o$ the #ean #otor unit potential )M>P+ siCe. /2 5his #ean M>P is calculated b( avera4in4 a nu#ber o$ individual M>Ps that have been sa#pled usin4 one o$ a variet( o$ approaches./- 5o obtain a representative #ean M>P, a si4ni$icant proportion o$ all M>Ps have to be sa#pled. A lar4er M>P

sub4roups based on the ?** scores. ?i$teen patients )613+ had an ?** score K8 and ,ere considered severel( $ati4ued. 5he #ean a4e o$ the severel( $ati4ued patients ,as 8A N -2 (ears, the nonOseverel( $ati4ued patients ,as 81 N -- (ears, and the health( controls ,as 87 N -6 (ears. 5hese / sub4roups o$ patients did not di$$er re4ardin4 distribution in a4e, se0, and residual de$icits and disabilit( )table -+. %ne patient had an ?-AD* score P1. 5his patient belon4ed to the nonQseverel( $ati4ued patient 4roup. ;one o$ the patients had clinical andRor electroph(s' iolo4ic si4ns o$ carpal tunnel s(ndro#e. 5he #edian M>;E o$ all GB* patients ,as -7- )9H@ -2<'/<<+ and lo,er than the M>;E o$ health( controls at 681 )/78'7-:+ (p M 2.22-+. ?urther#ore, the M>;E di$$ered bet,een the severel( $ati4ued patients and the non'severel( $ati4ued GB* patients )-/8 S9H@ :8'-7-T and /81 S-/2'678T,

1828

Neurolo y #$ November $;/ %2$&

Table $ 9emo raphy and residual clinical scores of '(S patients and healthy sub<ects
Non-severely at!gued pat!ents Severely at!gued pat!ents Control su"#e$ts %n & '() %n & *+) Age- y Se.- /01 Years s!n$e GBS d!agnos!s GBS d!sa"!l!ty s$ore /RC sum s$ore O2SS
81 )--+ -6 -1.: ):.2+ - )2'/+ :2 )8A':2+ - )2'/+ 8A )-2+ <1 1.< ):.<+ / )-'/+ 8A )8-':2+ / )2'8+

%n & -7+
88 )-/+ <<

p ,alue
;* ;* ;* ;* ;* ;*

+bbreviations. '(S 3 'uillain"(arre syndrome6 M-C 3 Medical -esearch Council6 NS 3 not si nificant6 =9SS 3 overall disability sum score! 9ata are presented as means (S9) or medians (inter0uartile ran e)! The p value is calculated between non" severely and severely fati ued patients! NS. statistically not si nificant (p a 2!2*)! '(S disability score ran es from 2 (>healthy?) to 1 (>dead@)! M-C sum score ran es from 2 (paralysis) to 12 (normal stren th)! =9SS ran es from 2 (?no si ns of disability?) to $% (?most severe disability?)!

respectivel(& p U 2.226+, and bet,een the / patient 4roups and the health( controls )$i4ure+. 5he M>;E in GB* patients ,as correlated to the ?** )p U 2.72, p = 2.2-+, to a4e )p U Q 2.6-,RP U 2.26+, the residual M@C su# score )p U 2.:< ,p M 2.22-+, GB* disabilit( score )p U Q2.6:, p U 2.26+, and the %D** )p U Q2 .46, p U 2.227+. Althou4h the M>;E o$ all GB* patients to4ether correlated ,ith a4e and clinical scores, there ,ere no di$$erences in a4e and clinical scores )M@C su# score, GB* disabilit( score, and %D**+ bet,een the / patient 4roups )table -+. *tep,ise linear re4ression, includin4 M>;E, a4e, and

and the dots represent the outliers of the MUNE of the severely fati ued '(S patients (n 3 $*)/ the non" severely fati ued '(S patients (n 3 %,)/ and the healthy sub<ects (n 3 $,)! '(S 3 'uillain"(arre syndrome6 MUNE 3 motor unit number estimate!

7i ure MUNE of '(S patients and healthy sub<ects

M@C su# score revealed M>;E as the onl( predictin4 variable )P Q2.6:, p U 2.2/+ $or the ?**. Besides M>;E, the #ean M>P siCe ,as also di$' $erent bet,een the / patient 4roups 8A V0V )9H@ 76' :7 C xD) $or the severel( $ati4ued GB* patients and /A V0V )9H@ -A'77 VaV+ $or patients ,ho ,ere not severel( $ati4ued (p U 2.2-+. 5he #ean M>P siCe ,as correlated to the ?** )p U Q2.68, p 2.26+. Conventional ;C* sho,ed no di$$erence in #otor nerve variables )distal #otor latenc(, distal CMAP, pro0i#al CMAP, #otor nerve conduction velocit(, and ?',aves+ bet,een the severel( $ati4ued and non' severel( $ati4ued patients )table e'- on the Neurology Ieb site at ,,,.neurolo4(.or4+. Co#pared ,ith the non'severel( $ati4ued patients, the severel( $ati4ued patients had lo,er *;AP a#plitudes o$ the #edian nerve )#edian -< V0V and A V0V, respectivel(& p U 2.2-+ and ulnar nerve )#edian -7 S0V and 1 !0V, respectivel(& p U 2.26+ as ,ell as lo,er sural sensor( conduction velocit(. *tep,ise linear re4ression )includin4 M>;E, #edian *;AP a#plitude, and ulnar *;AP a#plitude+ identi$ied M>;E as the onl( predictor $or ?** score.
DISC SSI#! 9n this stud(, ,e $ound that severe $ati4ue

''''','''''''''''''''''''''''','''''''''''''''''''''''',''''''''
Severely fati ued Notseverely fati ued 8ealthycontrol

The bottom and top of the box represent the %*th and A*th percentile/ the band represents the median value/ the whisBers represent the $2th and ;2th percentile/

in GB* is related to a0onal loss. A0onal loss ,as detected b( M>;E and not b( clinical e0a#ination and conventional neuroph(siolo4ic #easure#ents such as CMAP a#plitudes. *everel( $ati4ued GB* patients have, on avera4e, lo,er M>;Es than non' severel( $ati4ued GB* patients. Because the / patient 4roups did not di$$er re4ardin4 a4e and clinical scores, the di$$erence in M>;E #ust be e0plained other,vise. A lo, M>;E is an indication o$ pronounced a0onal

Neurology 81 November 19, 2013

loss. Hence, this stud( provides the $irst direct evidence $or a ph(siolo4ic basis underl(in4 e0perienced $ati4ue. *trikin4l(, the CMAP a#plitudes did not di$$er bet,een the / patient 4roups. 5his con$ir#s our

1830

Neurology 81 November 19, 2013

assu#ption that conventional electroph(siolo4ic #eth' ods and clinical e0a#ination are not su$$iciend( sensi' tive to detect the presence o$ #otor a0onal loss, especiall( ,hen reinnervation has occurred )resultin4 in preservation o$ the CMAP a#plitude+. A reduced M>;E in co#bination ,ith a nor#al CMAP a#plitude i#plies that in the severel( $ati4ued patients, the re#ainin4 M>s are lar4er )the sa#e potential is pro' duced b( $e,er M>s+, a $indin4 that is consistent ,ith a previous stud( o$ M> siCe in patients ,ith GB*. /8 9ndeed, anal(sis o$ the avera4e #ean M>P siCe reveals a lar4er #ean M>P siCe in the severel( $ati4ued 4roup. 5he =uestion o$ ho, this a0onal loss #a( lead to $ati4ue re#ains unans,ered. ;or#all(, upon voluntar( activation o$ a #uscle, the $irst M>s that are recruited are the s#allest ones./: 5hese allo, $or precise, $ine #ove#ents. 9$ #ore $orce is re=uired, lar4er M>s are recruited. Possibl(, this orderl( recruit#ent #echanis# has beco#e disturbed in severel( $ati4ued patients ,ith GB*, because reinnervation #ade previousl( s#all M>s lar4e. Earl( recruit#ent o$ these lar4e M>s #a( then easil( result in an overshoot o$ $orce. 9t is conceiv' able that this overshoot either direcd( or throu4h co#' pensator( processes results in increased $ati4ue. ?urther#ore, because a$ter nerve in!ur( the siCe' ordered or4aniCation o$ M> properties can be restored a$ter a period o$ ti#e, lar4e M>s that co#e $ro# collateral reinnervation $ire $ast and $ati4ue =uickl(, si#ilar to nor#all( lar4e M>s. 5his possibl( results in an increased clinical $ati4ue./< 9n this stud(, the severel( $ati4ued patients had lo,er *;AP a#plitudes than the non'severel( $ati4ued patients, also a si4n o$ a0onal loss. 9n addition, the sural nerve conduction velocit( ,as lo,er in the severel( $ati4ued patients. 5hese di$$erences #i4ht be attributable to chance caused b( #ultiple testin4. *tep' ,ise linear re4ression anal(sis sho,ed that the #edian and ulnar *;AP a#plitudes ,ere not predictors o$ the ?** score. 9t is likel( that the( re$lect the a0on loss in 4eneral and corroborate the results o$ the M>;E and #otor $iber loss. A possible li#itation o$ this stud( #i4ht be that the presence o$ $ati4ue can be in$luenced b( #an( $actors. Althou4h ,e e0cluded patients ,ith kno,n diseases that #i4ht in$luence $ati4ue, ,e did not screen $or sub' clinical diseases )such as h(poth(roidis# and ane#ia+. 5o support our $indin4s and h(pothesis, lon4itu' dinal studies in patients ,ith GB* are re=uired as ,ell as a si#ilar stud( in patients ,ith other peripheral neuropathies such as chronic in$la##ator( de#(elin' atin4 pol(neuropath(. 5his stud( provides evidence that residual #otor and sensor( a0onal loss is present in a substantial pro' portion o$ patients ,ho recovered ,ell clinicall( $ro# a previous episode o$ GB* and that it is associated ,ith $ati4ue in these patients. Althou4h the #echanis#s responsible $or this $ati4ue are still unkno,n, these re' sults provide a use$ul step to,ard their unravelin4.

A $H#R C#!$RIB $I#!S

Judith Drenthen stud( desi4n, data collection, anal(sis and interpretation, and dra$tin4 #anuscript. Bart C. Jacobs stud( desi4n, data interpretation, and supervision ,ritin4 #anuscript. Ellen M. Maathuis data collection, anal(sis and interpretation, and #anuscript revision. Pieter A. van Doo# and Gerhard H. Visser data interpretation, and #anuscript revision. Joleen Blok stud( desi4n, data anal(sis and interpretation, and supervision ,ritin4 #anuscript.

S$ D- F !DI!G
*upported b( the Prinses Beatri0 ?onds )research 4rant IA@2:'/-8+.

DISC,#S R"
J. Drenthen has nothin4 to disclose. B. Jacobs has received $undin4 $or travel $ro# Ba0ter 9nternational 9nc., and has received research $undin4 $ro# the ;etherlands %r4aniCation $or Health @esearch and Develop#ent, Eras#us MC, the Prinses Beatri0 *pier$onds, and the GB*'C9DP ?oundation 9nternational. E. Maathuis has nothin4 to disclose. P. van Doo# has received $undin4 $or travel $ro# Ba0ter 9nternational 9nc., and has received research $undin4 $ro# the ;etherlands %r4aniCation $or Health @esearch and Develop#ent, the Prinses Beatri0 *pier$onds, and the GB*'C9DP ?oundation 9nternational. He received honoraria $or participation in steerin4 co##ittees o$ the 5alecris 9CE *tud( and %ctaphar#a. G. Visser has nothin4 to disclose. J. Blok received a research 4rant $ro# the Prinses Beatri0 *pier$onds )IA@2:'/-8+. Go to ;eurolo4(.or4 $or $ull disclosures.

Received May 17, 2013. Accepted in final form August 21, 2013. REFERENCES
-. van Doorn PA, @ues L, Jacobs BC. Clinical $eatures, patho4enesis, and treat#ent o$ Guillain'BarrK s(ndro#e. Lancet ;eurol /221&< A6A'A82. /. Hu4hes @AW *,an AV, @aphael JC, Annane D, van Gonin4sveld @, van Doorn PA. 9##unotherap( $or Guillain' Barre s(ndro#e a s(ste#atic revie,. Brain /22<& -62 //78'//8<. 6. Merkies 9*, ?aber CG. ?ati4ue in i##une'#ediated neuropathies. 6):S2Q3~S207. 7. de Vries JM, Ha4e#ans ML, Buss#ann JB, van der Ploe4 A5, van Doo# PA. ?ati4ue in neuro#uscular disor' ders $ocus on Guillain'Barre s(ndro#e and Po#pe disease. Cell Mol Li$e *ci /2-2&:< <2-'<-6. 8. Merkies 9*, *ch#itC P9, *a#i!n JP, van der Meche ?G, van Cause Doorn and PA ?ati4ue in i##une'#ediated ;europath( ;eurolo4( Group. pol(neuropathies. -AAA&86 -:71'-:87. :. Galk#an J*, *chillin4s ML, X,arts MJ, van En4elen BG, Blei!enber4 G. 5he develop#ent o$ a #odel o$ $ati4ue in neuro#uscular disorders a lon4itudinal stud(. J Ps(choso# @es /22<&:/ 8<-'8<A' <. Garssen MP, van Doorn PA, Visser GH. ;erve conduc' tion studies in relation to residual $ati4ue in Guillain'Barre s(ndro#e. J ;eurol /22:&/86 18-'18:. 1. MartineC'?i4ueroa A, Hansen *, Bailant(ne JP. A =uantitative electroph(siolo4ical stud( o$ acute idiopathic pol(neuritis. J ;eurol ;eurosur4 Ps(chiatr( -A<<&72 -8:' -:-. A. van Gonin4sveld @, *ch#itC P9, Meche ?G, et al. E$$ect o$ #eth(lprednisolone ,hen added to standard treat#ent ,ith intravenous i##uno4lobulin $or Guillain'Barre s(n' dro#e rando#ised trial. Lancet /227&6:6 -A/'-A:. -2.van der MechK ?G, *ch#itC P9. A rando#iCed trial co#' parin4 intravenous i##une 4lobulin and plas#a e0chan4e European 9n$la##ator( ;euro#uscul Disord /2-/&//)suppl

5reat#ent

)9;CA5+