nervebiopsy and conduction studies in neuropathywere absent or weak, knee jerks and tendon jerks in...

Journal ofNeurology, Neurosurgery, andPsychiatry, 1977, 40, 1072-1082

Nerve biopsy and conduction studies in diabeticneuropathyFRIEDRICH BEHSE, FRITZ BUCHTHAL, ANDFRITS CARLSEN

From the Laboratory of Clinical Neurophysiology, Rigshospitalet, the Institute of Neurophysiology,and Physical Laboratory II, University of Copenhagen, Denmark

SUMMARY Morphological findings in sural nerves were related to nerve conduction in 12patients with diabetic neuropathy, five with mainly sensory involvement, four with severe,symme,trical sensory-motor polyneuropathy, and three with multiple mononeuropathy. All hadloss of large and small myelinated and of unmyelinated fibres, even early in the disease;segmental remyelination was the most prominent myelin alteration in teased fibres, segmentaldemyelination was found in only a few fibres. Axonal degeneration and Schwann cell damageseem to proceed independently of each other. The relation between recorded conduction velocityand that expected from the diameter of the largest fibres indicated that slowing of 20 to 30%was due to causes other than fibre loss; a grossly diminished conduction velocity was causedmainly by fibre loss. Electrophysiological findings in the sural nerve were largely representativeof findings in other nerves, though abnormalities were less marked in the median nerve. In halfthe endoneurial vessels from diabetic neuropathy the perivascular space was thickened or con-tained more layers of basal laminae than normal. The same abnormalities were found inone-quarter of the endoneurial vessels from other acquired neuropathies.

The slowing in conduction which characterisesmost sensory and motor nerves in diabetic neuro-pathy (Lamontagne and Buchthal, 1970) has beenconsidered to be due to segmental demyelination(Gilliatt, 1966; Thomas and Lascelles, 1966; Chopraet al., 1969). The use of electron microscopy andteased fibre preparations have led to conflictingresults and explanations: is demyelination oraxonal loss the primary and main lesion, or doeseach develop independently of the other?To clarify this problem, we have quantitated

morphological findings in sural nerves and relatedthem to abnormalities in conduction in patientswith diabetic neuropathy.

Patients

Twelve consecutive patients with diabetes mellitus,abnormal glucose tolerance tests, and signs andsymptoms of neuropathy were examined (Table).Near relatives of six patients had diabetes mel-litus. In lI the diabetes began in maturity (>40Address for reprint requests: Professor F. Buchthal, Laboratory ofClinical Neurophysiology, Rigshospitalet, Copenhagen, Denmark.Accepted 2 June 1977

years). In four patients the metabolic disturbancewas severe and in three of these poorly regulated.Other manifestations of diabetes were retinopathy(five patients), perforating ulcer of the leg (onepatient), or other trophic changes (two patients).Four patients had hypertension, and none hadsigns of nephropathy or liver disease.There was no evidence of other causes of the

neuropathy than diabetes. In seven patients neuro-logical signs and symptoms preceded or coincidedwith the diagnosis of diabetes mellitus. In fivepatients signs and symptoms of neuropathy weremainly sensory (diminished sensitivity to touch,pinprick, and vibration), the only clinical motorinvolvement being weakness and wasting of theextensor digitorum brevis muscle. Four patientshad severe symmetrical weakness and wastingmainly in distal muscles of the lower limbs in ad-dition to sensory impairment. In three patientsweakness and wasting were localised to proximalor distal muscles of one leg; sensory impairmentwas bilateral in two. In all patients ankle jerkswere absent or weak, knee jerks and tendon jerksin the arms were absent as well, with the ex-ception of the five patients with mainly sensory

1072

Protected by copyright.

on June 17, 2021 by guest.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.40.11.1072 on 1 N

ovember 1977. D

ownloaded from

http://jnnp.bmj.com/

Nerve biopsy and conduction studies in diabetic neuropathy

Table Clinical data on 12 patients with diabetic neuropathy

Patient Age Sex Diabetes mellitus Duration Sensory impairment Weakness(yr) Duration Severity Treatment ofneuropathy Feet Hands

Sensory neuropathy98 58 M 14 years severe insulin I year + + none124 51 F recently mild diet 1j years + 0 none191 56 F 2 years mild diet 2 years + 0 none229 55 M 5 months mild diet 5 months + 0 none240 43 F 23 years severe insulin I year + + none

Symmetrical sensory-motor neuropathy28 71 F I year severe insulin 6 months + 0 prox. + dist.166 51 F 5 years mild oral 2 years - distal198 67 F 1 year mild diet 1 year ++ distal228 71 M > 14 years mild diet 14 years + + distal

Multiple mononeuropathy81 58 F 15 years mild oral 6 months (+) 0 left quadr.

222 70 M 12 years severe insulin 1 year -V 0 left leg225 55 M recently mild diet 8 months + 0 left ant. tib.

involvement. All 12 patients complained of paraes-thesia, and nine complained of severe pain in thelegs. The spinal fluid protein was normal.Seven patients were re-examined one to eight

years after the first study. One of three patientswith severe sensory-motor involvement had im-proved and would be classified as mainly sensoryneuropathy at the second study; conductionvelocity was still severely slowed. The neuropathyof the other six patients (two with symmetricalsensory-motor, two with mainly sensory neuro-pathy, and two with multiple mononeuropathy)had progressed slightly with respect to sensoryand improved in five with respect to motor im-pairment; they would still be classified as at thefirst examination; conduction velocities were aswhen the patients were first seen, and amplitudesof the sensory potentials had diminished further.In one patient with multiple mononeuropathy aparalysis of the distal segment of the mediannerve had developed suddenly, simulating acarpal tunnel syndrome, similar to the patient ofGilliatt and Willison (1962). There was, however,no pain.

Method

SENSORY AND MOTOR NERVE CONDUCTION (Behseand Buchthal, 1971).Sensory conduction was determined along thesural nerve (lateral malleolus to 120 mm proximalto it or to the popliteal fossa), along the super-ficial peroneal nerve (from the superior retina-culum at the ankle to the capitulum fibulae),along the saphenous nerve (from the medialepicondyle to the inguinal ligament), along thedistal portion of the posterior tibial nerve (fromtoe I to the medial malleolus), and along the

distal segments of the median nerve (from digitsI and III to the wrist). Sensory nerve actionpotentials which were less than 3 ,uV were re-corded by electronic averaging of 500-2000potentials (Buchthal and Rosenfalck, 1966).Motor nerve conduction and distal and proximal

latencies were determined along the deepperoneal, the femoral, the posterior tibial, andthe median nerves.

ELECTROMYOGRAPHYElectromyography was performed in two to fivedistal and proximal muscles in the lower andupper limbs.

NERVE BIOPSYThirty millimetres of the sural nerve were re-moved in toto, just proximal to the lateralmalleolus, and fixed in 2% buffered isotonicglutaraldehyde (36 h). For the preparation ofteased fibres, a 10 mm length was postfixed in 1%buffered osmium tetroxide (4 h) and stored in0.2 M sucrose. For light and electron microscopythe specimen was post-fixed in 1% bufferedosmium tetroxide (2 h), dehydrated in gradedconcentrations of ethanol and embedded in Epon812 (Behse et al., 1974). Transverse sections(3-5 jum thick) for light microscopy were stainedwith 1% paraphenylene-diamine (Holliinder andVaaland, 1968), and ultrathin sections for electronmicroscopy were stained with uranyl acetate andlead citrate (Reynolds, 1963).Light microscopy In each biopsy specimen thetransverse endoneurial area was measured, andin an area of 0.4-0.6 mm2, sampled from allfascicles, the total number and size distributionof myelinated fibres and the number of groups of

D

1073



j.com/

J Neurol N


ovember 1977. D

ownloaded from


Friedrich Behse, Fritz Buchthal, and Frits Carlsen

three or more regenerating fibres ('clusters') weredetermined (Behse et al., 1972, 1974).Electron microscopy An area of 10 000-20 000 jUm2, sampled from three different fascicles,was analysed at a final magnification of X10 000-12 000. The incidence was determined of fibreswith abnormalities in the fine structure of theaxon or of the myelin sheath, of bands ofBungner, and of onion-bulb formations. Unmy-elinated fibres were identified by criteria describedelsewhere (Behse et al., 1975); their number andsize distribution, the incidence of degeneratingfibres, characterised by loss of axonal organelles,and the number of Schwann cell subunits contain-ing or devoid of unmyelinated fibres weredetermined.

Finally, we measured the thickness of the peri-vascular space of the endoneurial vessels (finalmagnification X10000-20000). To avoid errorsdue to oblique sectioning, the smallest thicknesswas measured from the outer membrane of theendothelial cells to the outer concentric layer ofbasal laminae. Measurements across the nuclearregion of pericytes were avoided. We did notmeasure the thickness of the individual layer ofbasal laminae (Bischoff, 1968).

Findings from electron microscopy were com-pared with those in six nerves from controls and,with respect to endoneurial vessels, with findingsin nerves from 13 patients with other types ofacquired neuropathy.Teased fibres In nine nerves 34 to 73 myelinatedfibres, more than 7 tm in diameter, were teasedin 60% glycerol (Vizoso and Young, 1948). Theincidence of the following abnormalities was deter-mined from micrographs at a magnification ofX 160-200 for measurement of internodal length,and of X800 for measurement of diameter(Behse and Buchthal, 1977b): (a) segmental de-myelination; (b) widening of the nodal gap tomore than 10 ,um and paranodal demyelinationover a length of up to 300 um; (c) fibres whichhad internodes of normal length, containing someinternodes with an abnormally thin myelin sheath;(d) segmental remyelination; two or more short,often thinly myelinated segments between seg-ments of normal length; (e) solitary intercalatedsegments; (f) regenerated fibres-all internodalsegments were short (0.2-0.6 mm) and had adisproportionately thick diameter.

OTHER TESTSGlucose in blood and urine, glucose tolerancetest, liver function tests, and investigation of thecerebrospinal fluid were performed by standardtechniques.

Results

NERVE CONDUCTIONIn 76% of 25 sensory nerves examined in thelower extremities, and in 30% of 20 branches ofmedian nerves, conduction velocity was slowed.In motor nerves slowing in conduction occurredas often along the nerves of the lower as alongthose of the upper extremities. Sensory conductionvelocity was normal or slightly slowed in patientswith only sensory involvement and in patients withmultiple mononeuropathy; in two patients withsevere sensory and motor involvement, slowingwas marked (to less than 60% of normal, that is,less than 30-35 m/s). Slowing of motor conduc-tion also tended to be more marked in patientswith sensory-motor involvement (Fig. 1).

11 0

100z0

0z

0

o 70

Z 50Un

I

6

30k-

20k

_I

Ax'I

1 0k

0 SENSORY TYPE I |SENS-MOTOR TYPE|MULTI PLE

IMONONE UR.1Fig. 1 Sensory and motor nerve conduction inpatients with the three types of diabetic neuropathy.Conduction velocity is given as a percentage of thenormal average, matched for age; motor latencies aregiven as reciprocal latencies. Each vertical lineconnects findings in different nerves in the individualpatient; ---- =lower 95% confidence limit of normtal.Symbols: Sensory nerves: sural (X), superficialperoneal (0), posterior tibial (O), saphenous (u),median (0, digits I and III). Motor nerves, velocity:deep peroneal (0). Motor latencies: m. tibialisanterior and m. peroneus longus (v), m. extensordigitorum brevis (-), m. flexor hallucis brevis (*),m. quadriceps (i), m. abductor pollicis brevis (i).

1074

T-t

i40 _

XX



j.com/

J Neurol N


ovember 1977. D

ownloaded from



AMPLITUDE OF THE SENSORY ACTION POTENTIALIn 65% of 48 nerves examined in the upper andlower extremities the amplitude of the sensorynerve action potentials was diminished. Abouthalf the potentials were split up into many com-ponents. The reduction in amplitude was mostmarked in patients with severe sensory and motorinvolvement (amplitudes 7 ,um) and small(


mm2

2.5-

2.0-

1.5-

1.0-

0.5-

O O-

ENDONEUR.AREA

2 C IDN

XXx

IXXX

0

o a

0

BOO

7_

00.

9000-

7000-

5000- Xx

3000-

1000-

0-

ALLF IBRES

C DNx

x

xx ox

xxx

0

03000a

no.7000-

5000-

3000-

1000-

FIBRES< 7pmC IODN

x

xx

xxx xx x

x

0

0 0

00ao

no.3500-

2500-

1500-

500-

0-

FIBRES>7pmC IODNx

xxxxxxXX

x

x

00

0

00 001 0

1.0a

amE-

no.220-

120 -

80 -

CLUSTERS

C DN

40 1 X

0-

X X

XexlXXI

- O-o

X

00

00

0

LaU-

Fig. 3 Endoneurial area,number of myelinated fibres andof clusters of regenerating fibresin sural nerves from patientswith diabetic neuropathy (DN)and in controls (C,X). Symbolsfor the different types ofdiabetic neuropathy as in Fig. 2.

fibres, evidenced by the presence of clusters, wasseen in nine nerves.

In nerves of patients with sensory neuropathy(Fig. 4, right) and with multiple mononeuropathy(Fig. 4, left), the histogram of fibre diameterswas bimodal, and in the four nerves from patientswith sensory-motor neuropathy the histogram wasunimodal or unimodal and skew.Conduction velocity as related to fibre diameterConduction velocity and the diameter of my-elinated fibres are proportional and maximum con-duction velocity can be calculated from the diam-eter of the largest fibres by the conversionfactor (Gasser and Erlanger, 1927), found in thehuman sural nerve to be 4.3-+0.1 (Buchthal et al.,1975). In all diabetic nerves except one the con-duction velocity was lower than to be expectedfrom the fibre diameter (Fig. 5). In six nerves itdeviated at most 20%, exceeding slightly thelargest deviation observed in nerves from controls.In five nerves the recorded velocity was 20-30%slower than to be expected from the fibre diam-eter, the additional slowing being relatively thesame whether the recorded velocity was 15 m/s or42 m/s.Teased fibres One or the other abnormality in themyelin sheath was found in every nerve; abnor-malities were present in 35-80% of the fibres, ascompared with at most 30% in control nerves. Asto the type of abnormality (Fig. 6), fibres with de-myelinated segments were found in the nerves ofthose two patients with neuropathy of the shortestduration; in all it occurred in only eight of 502

teased fibres. Paranodal demyelination was alsoscarce, occurring in three nerves, in all in 17 fibres.Segmental remyelination was the most frequentabnormality; it was seen in all but one nerve andoccurred in 20-40% of teased fibres. Usually onesingle internodal segment per teased fibre was re-placed by two to four short remyelinated segments.Other abnormalities of the myelin sheath (solitaryintercalated segments and fibres which had somesegments with a disproportionately thin myelinsheath), though present in most nerves, occurredin only a small proportion of teased fibres. Solitaryintercalated segments were usually found at asingle site of teased fibres. One-sixth of the fibreswith segmental remyelination had in addition oneor more solitary intercalated segments. Finally,few nerves contained some fibres with abnormalRanvier nodes, probably indicating local repair(Thomas and Lascelles, 1966). Only two nerveshad a marked number of regenerated fibres.Conduction velocity related to myelin abnormalitiesIn the nine nerves from patients with diabeticneuropathy, the incidence of abnormalities in themyelin sheath was not related to conductionvelocity, whether the proportion of abnormalitieswas taken of the number of teased fibres or of thenumber of internodal segments. However, when91 nerves were pooled (including those from dia-betic neuropathy), conduction velocity decreasedwith increasing incidence of myelin abnormalities(P

12(

10(cn

( 8CiLU-

-J 6ClL

2U-0

I

CDz O-J-J< 0C]C

- n

0.0L L

--3

+

I L_7 8 9 10 11 12 13 14

DIAMETER, pm

0~~~~i, .,D


SEGMENTAL PARANODAL THIN MYELIN,DEMYELINATION DEMYELINATION NORMAL INTER-

SEGMENTAL SOLITARY INTER- REGEN.FIBRESREMYELINATION CAL.SEGMENT(S)

50 1-r -ll l ll

20 40 m/s 60MAXIMUM VELOCITYPREDICTED FROM DIAMETER

Fig. 5 Maximum conduction velocity along suralnerve determined from the sensory action potential, asa function of the velocity predicted from diameter ofthe largest myelinated fibres. Full line indicatesequality between recorded and predicted velocity,dashed line represents a 20% slower recorded thanpredicted velocity. Symbols for the different types ofdiabetic neuropathy as in Fig. 2.

lished data). The large scatter in the relation be-tween myelin damage and conduction velocitymay be due in part to the fact that conductionvelocity was determined over a length of 120 mm,whereas abnormalities in teased fibres were deter-mined over at most 10 mm. When abnormalitieswere counted over half the length of the teasedfibre, their incidence decreased by about one-third.The incidence of abnormalities would probablyincrease if a representative number of fibres couldhave been teased over, for example, 20 mm.Electron microscopy of myelinated nerve fibresThe cross-sectional area investigated contained542 myelinated fibres. Only two fibres of one nerveshowed loss of axonal organelles, indicating thebeginning of axonal degeneration. Signs of demy-elination, present in two nerves, were found inless than 1%; signs of incipient remyelination (dis-proportionately thin myelin sheath) were presentin only three nerves, in 2% of the fibres examined.Otherwise, myelin thickness was normal relativeto the axonal diameter.Unmyelinated fibres Unmyelinated fibres andtheir Schwann cells were assessed in 10 nerves.The most severe abnormalities were found in theseven nerves from patients with sensory or sensory-

4 0 -

3 0 -

2 0 -

1 0 -

x

x

xxxx xX X

0

0 00N0

0 0

0x

xxxx x

Fig. 6 Incidence of fibres with myelin abnormalitiesand of regenerated fibres among teased fibres fromsural nerves of patients with diabetic neuropathy, andfrom controls (X). Symbols for the different types ofdiabetic neuropathy as in Fig. 2.

motor neuropathy: a diminished number of un-myelinated fibres (three nerves), an increasednumber of fibres undergoing degeneration (fournerves), and a markedly increased number ofSchwann cell subunits devoid of axons (six nerves).Two of the three nerves from patients with mul-tiple mononeuropathy had one or the other ab-normality in unmyelinated fibres or their Schwanncells, though abnormalities were mild. In all 10nerves, the mean diameter and the size distribu-tion of unmyelinated fibres were normal.Endoneurial vessels The thickness of the peri-vascular space and the number of layers of basallaminae (Fig. 7) were determined in nine nerves ofpatients with diabetic neuropathy, and comparedwith seven controls and with 13 nerves of patientswith other types of acquired neuropathy. Becauseof indistinct separation of the individual layers(absence of connective tissue between them), theirnumber could not be counted in six vessels frompatients with diabetic neuropathy and in twovessels from control subjects.

In control subjects, the thickness of the perivas-cular space ranged from 1-4 ,um, and it contained

m/s60

0wz

wZw4

cz Z40o uj4t- 0CL0>

z

7-

00

a

0

0

000000

0

0oU

x

X Xxxx

1078



j.com/

J Neurol N


ovember 1977. D

ownloaded from



,- /. 4 -;,

5.pm ..........

5pm

Fig. 7 Electron micrographsof endoneurial vessels. Left,from normal nerve; right, fromnerve of a patient with diabeticneuropathy. Note increasedthickness of pericapillary spaceand increased number ofconcentric layers of basallaminae interwoven withcollagen fibrils in diabeticneuropathy.

five to 10 concentric layers of basal laminae or offragments of basal laminae interwoven with layersof collagen fibrils (Fig. 7). In seven vessels fromnerves of patients with diabetic neuropathy thediameter of the perivascular space was increased(Fig. 8), and nine vessels contained 11-17 layersof basal laminae; in all, half the vessels showedeither one abnormality or the other, or both.Similar abnormalities were found in one-quarterof the vessels from nerves of patients with othertypes of acquired neuropathy.

Discussion

In most early studies, loss of myelinated fibreswas considered to be the main pathology ofdiabetic neuropathy (Greenbaum et al., 1964).Thomas and Lascelles (1965, 1966) found seg-mental demyelination in teased fibres of all nerves,whereas axonal loss occurred particularly insevere chronic cases. They considered it an openquestion whether the axonal loss was secondary tosegmental demyelination or occurred independ-ently. Since demyelination but not axonal loss wasfound in diabetics without neuropathy, Chopra etal. (1969) believed segmental demyelination to bethe primary abnormality. This confirmed the viewof Bischoff (1968) that a metabolic Schwann celllesion is the primary defect in diabetic neuropathy.However, neuropathy that occurred early in the

course of diabetes was associated with axonal de-generation (Bischoff, 1973). Thomas and Eliassen(1975), in their recent review, considered it un-settled whether demyelination is the primarylesion in diabetic neuropathy or is secondary toaxonal degeneration. Our findings are compatiblewith the assumption that demyelination andaxonal loss are independent processes.As to demyelination, the most frequent abnor-

mality in teased fibres was segmental remyelina-tion, though it was far less frequent than in thehypertrophic type of peroneal muscular atrophy(Behse and Buchthal, 1977a). Onion-bulb forma-tions were absent both on light and electronmicroscopy; they were found by Ballin andThomas (1968) in seven of 10 biopsies from patientswith diabetic neuropathy. Segmental remyelinationwas more frequent than in those neuropathies withaxonal loss as the main pathology (Behse andBuchthal, 1977a, b). Paranodal myelin damage isan early change both in axonal degeneration andin segmental demyelination (Ballin and Thomas,1969). When it is secondary to axonal degenera-tion, myelin damage tends to occur at multiplesites along certain fibres (Dyck et al., 1971). Inour material paranodal or segmental de- or re-myelination of multiple sites occurred in two ofnine nerves in half the teased fibres with myelinabnormalities. In contrast to findings of Bischoff(1973), we found de- or remyelination also in those

1079



j.com/

J Neurol N


ovember 1977. D

ownloaded from



6

4

2

0

Fig.

thic

pati13 rneu

patipat]

signfibrisho:dimsizeeiglpatiamj

pot(pot(nonmoI

logi

DIABETIC NEUROPATHY scribed by Williams and Mayer (1976), and were(n=24) as marked as in nerves of patients of the sensory

type of neuropathy.Brown et al. (1976) thought the pain in three

diabetic patients with sensory neuropathy to bedue to a disproportionate loss of small myelinatedfibres and a disproportionate increase in thenumber of small unmyelinated fibres. This 'small-fibre neuropathy' was not found in our patientswith sensory neuropathy who complained of severepain.

In most sural nerves, the conduction velocity_CONTROLS was 10-30% slower than to be expected from theL(n=23) diameter of the largest myelinated fibres. This

slowing was relatively the same whether thelargest fibres were lost or preserved. The additionalslowing was probably due in part to myelindamage. It is unsettled to what extent othercauses play a role-for example, abnormalities

I I without morphological concomitants as assumedin experimental diabetes (Sharma and Thomas,1974).

OTHER NE UROPATH IES The fact that many clinically normal muscles(n=24) in the mainly sensory type of neuropathy had

electromyographic abnormalities raises thequestion whether the clinical differences betweenthe sensory and the sensory-motor types reflect adifference in severity rather than in type. Fibreloss was more pronounced in patients with thesensory-motor than in those with the sensory type,

l } l as was the diminution in the amplitude of thesensory action potentials and the slowing in nerve

) 1 2 3 4 5 6 pm 7 conduction velocity. But the relation between the8 Number of endoneurial vessels with different recorded velocity and the velocity expected fromkness of the perivascular space in nine nerves of diameter of the largest fibres was the same, indi-ients with diabetic neuropathy, seven controls, and cating that there is no evidence for differentierves of patients with other types of acquired mechanisms that cause the abnormalities in con-ropathy. n=number of vessels. duction in the two types. Moreover, one of the

patients with the sensory-motor type would havebeen classified as having the sensory type one year

ients in whom signs and symptoms of neuro- after the first study.hy preceded the diagnosis of diabetes. Compared with nerves from control subjects,is to axonal degeneration, all nerves showed half the endoneurial vessels from diabetic neuro-is of loss of myelinated and unmyelinated pathy had thickening of the perivascular space ores, even in patients with neuropathy of the an increased number of concentric layers of basalrtest duration. Our findings agree with the laminae. Similar observations were reported byinished density of myelinated fibres of all Bischoff (1965, 1973), who found, in addition, ans, reported by Chopra and Hurwitz (1969) in increase in the thickness of the individual basalit of nine patients with sensory-motor neuro- laminae, which we did not measure in our patients.hy. Fibre loss was also evident from the low Some of the changes in vessels from diabetic nerveplitude of evoked nerve and muscle action may have been overrated, since the perivascularentials and from the loss of many motor unit space of endoneurial vessels has been assumed toentials in the electromyogram. In clinically be normally as thin as around intramuscularl-involved nerves of patients with multiple vessels (Garcin and Lapresle, 1968; Arne et al.,noneuropathy, histological and electrophysio- 1972; Vital et al., 1973). In fact, in controls theical abnormalities were similar to those de- perivascular space around endoneurial vessels was

no6

6

4

2

08

1080



j.com/

J Neurol N


ovember 1977. D

ownloaded from



thicker than around intramuscular vessels. Thesame abnormalities as in diabetic neuropathy werepresent in one-quarter of the endoneurial vesselsin other types of acquired neuropathies.

We are indebted to the Departments of Neuro-medicine, Neurosurgery, and Medicine (TA),Rigshospitalet, to the Departments of Neurologyof the Municipal Hospital and Bispebjerg Hospital,Copenhagen, the Frederiksborg County Hospital,Hiller0d, the Hvid0re Hospital for Diabetics,Copenhagen, and the Department of Rheuma-tology, Copenhagen County Hospital, Hvidovre,for permission to examine patients under theircare, and to the staff of the Department ofNeurosurgery, Rigshospitalet, Copenhagen, forperforming the sural nerve biopsies. The work wassupported by grants from the Muscular DystrophyAssociations of America Inc. and the MichaelsenFoundation, Copenhagen.

References

Arne, L., Vital, Cl., Vallat, J.-M., Yung. A., Leblanc,M., and Martin, F. (1972). Neuropathie peripheriquediabetique. Etudle anatomo-clinique avec ultrastruc-ture d'une biopsie nerveuse. Revue Neurologique,126, 115-126.

Ballin. R. H. M., and Thomas, P. K. (1968). Hyper-trophic changes in diabetic neuropathy. Acta Neuro-pathologica (Berlin), 11, 93-102.

Ballin. R. H. M., and Thomas, P. K. (1969). Changesat the nodes of Ranvier during Wallerian degenera-tion: an electron microscope study. A cta Neuro-pathologica (Berlin), 14, 237-249.

Behse. F.. and Buchthal, F. (1971). Normal sensoryconduction in the nerves of the leg in man. Journalof Neurology. Neurosurgery, and Psychiatry, 34,404-414.

Behse, F.. Buchthal, F.. Carlsen, F., and Knappeis,G. G. (1972). Hereditary neuropathy with liabilityto pressure palsies. Brain, 95, 777-794.

Behse, F., Buchthal. F., Carlsen, F., and Knappeis,G. G. (1974). Endoneurial space and its constituentsin the sural nerve of patients with polyneuropathy.Brain. 97, 773-784.

Behse, F., Buchthal, F.. Carlsen, F., and Knappeis,G. G. (1975). Unmyelinated fibres and Schwanncells of sural nerve in neuropathy. Brain, 98, 493-510.

Behse. F., and Buchthal. F. (1977a). Peroneal mus-cular atrophy (PMA) and related disorders. II.Histological findings in sural nerves. Brain, 100,67-85.

Behse, F.. and Buchthal, F. (1977b). Alcoholic neuro-pathy. Clinical. electrophysiological and biopsyfindings. A nnals of Neurology, 2, 95-1 10.

Bischoff, A. (1965). Die diabetische Neuropathie.Praxis. 54, 723-729.

Bischoff, A. (1968). Diabetische Neuropathie. Patholo-gische Anatomie, Pathophysiologie und Pathogeneseauf Grund elektronenmikroskopischer Untersuch-ungen. Deutsche Medizinische Wochenschrift, 93,237-241.

Bischoff, A. (1973). Ultrastructural pathology of per-ipheral nervous system in early diabetes. In Advancesin Metabolic Disorders. Suppl. 2, pp. 441-449.Edited by R. A. Camerini-Davalos and H. S. Cole.Academic Press: New York and London.

Brown, M. J., Martin. J. R., and Asbury, A. K. (1976).Painful diabetic neuropathy. A morphometric study.Archives of Neurology (Chicago), 33, 164-171.

Buchthal. F., and Rosenfalck, A. (1966). Evokedaction potentials and conduction velocity in humansensory nerves. Brain Research, 3, 1-122.

Buchthal, F., Rosenfalck, A., and Behse, F. (1975).Sensory potentials of normal and diseased nerves.In Peripheral Neuropathy, Volume 1, pp. 442-464.Edited by P. J. Dyck, P. K. Thomas. and E. H.Lambert. W. B. Saunders: Philadelphia.

Chopra, J. S., and Hurwitz, L. J. (1969). Sural nervemyelinated fibre density and size in diabetics. Jour-nal of Neurology, Neurosurgery, and Psychiatry, 32,149-154.

Chopra. J. S.. Hurwitz. L. J., and Montgomery,D. A. D. (1969). The pathogenesis of sural nervechanges in diabetes mellitus. Brain, 92, 391-418.

Dyck, P. J., Johnson, W. J., Lambert, E. H.. andO'Brien, P. C. (1971). Segmental demyelinationsecondary to axonal degeneration in uremic neuro-pathy. Mayo Clinic Proceedings, 46, 400-431.

Fullerton, P. M., Gilliatt. R. W.. Lascelles, R. G.,and Morgan-Hughes. J. A. (1965). The relation be-tween fibre diameter and internodal length inchronic neuropathy. Journal of Physiology, 178,26-28P.

Garcin. R., and Lapresle. J. (1968). La micro-angiopathie dlu nerf peripherique au cours du diabetesucre. Etude ultrastructurale de biopsies nerveuseset musculaires prelevees dans trois cas de neuro-pathie diab6tique. Bulletins et Memoires de laSociete Medicale des H6pitaux de Paris. 119, 363-381.

Gasser, H. S.. and Erlanger, J. (1927). The role playedby the size of the constituent fibres of a nervetrunk in determining the form of its action poten-tial wave. American Journal of Physiology, 80, 522-547.

Gilliatt, R. W. (1966). Diabetic neuropathy. In Dia-betes mellituis. pp. 64-73. Edited by L. J. P. Duncan.University Press: Edinburgh.

Gilliatt. R. W., and Willison. R. G. (1962). Peripheralnerve conduction in diabetic neuropathy. Journalof Neurology, Neurosurgery. and Psychiatry, 25,11-18.

Greenbaum. D., Richardson. P. C.. Salmon, M. W..and Urich, H. (1964). Pathological observations onsix cases of diabetic neuropathy. Brain. 87, 201-214.

Hollander, H., and Vaaland, J. L. (1968). A reliablestaining method for semi-thin sections in experi-mental neuroanatomv. Brain Research. 10, 120-126.

1081



j.com/

J Neurol N


ovember 1977. D

ownloaded from


1082

Lamontagne, A., and Buchthal, F. (1970). Electro-physiological study in diabetic neuropathy. Journalof Neurology, Neurosurgery, and Psychiatry, 33,442-452.

Reynolds, E. S. (1963). The use of lead citrate athigh pH as an electron-opaque stain in electronmicroscopy. Journal of Cell Biology, 17, 208-212.

Sharma, A. K., and Thomas, P. K. (1974). Peripheralnerve structure and function in experimental dia-betes. Journal of the Neurological Sciences, 23,1-15.

Thomas, P. K., and Lascelles, R. G. (1965). Schwann-cell abnormalities in diabetic neuropathy. Lancet, 1,1355-1357.

Thomas, P. K., and Lascelles, R. G. (1966). The path-ology of diabetic neuropathy. Quarterly Journal ofMedicine, NS, 35, 489-509.


Thomas, P. K., and Eliassen, S. G. (1975). Sensorypotentials of normal and diseased nerves. In Per-ipheral Neuropathy. Volume 2, pp. 956-981. Editedby P. J. Dyck, P. K. Thomas, and E. H. Lambert.W. B. Saunders: Philadelphia.

Vital, Cl., Vallet, J. M., Leblanc, M., Martin, F., andCoquet, M. (1973). Les neuropathies p6ripheriquesdu diabete sucre. itude ultrastructurale de 12-casbiopsies. Journal of the Neurological Sciences, 18,381-398.

Vizoso, A. D., and Young, J. Z. (1948). Internodelength and fibre diameter in developing and re-generating nerves. Journal of Anatomy (London),82, 110-134.

Williams, J. R., and Mayer, R. F. (1976). Subacuteproximal diabetic neuropathy. Neurology (Minne-apolis), 26, 108-116.



j.com/

J Neurol N


ovember 1977. D

ownloaded from

nervebiopsy and conduction studies in neuropathywere absent or weak, knee jerks and tendon jerks in...

Documents