peripheral disease - journal of neurology, …jnnp.bmj.com/content/jnnp/35/1/22.full.pdf ·...

Journal of Neurology, Neurosurgery, and Psychiatry, 1972, 35, 22-30

Peripheral neuropathy in chronic liver disease:clinical, electrodiagnostic, and nerve biopsy findings

R. P. KNILL-JONES, C. J. GOODWILL, A. D. DAYAN,AND ROGER WILLIAMS

From the Medical Research Council Group on Metabolism and Haemodynamics of Liver Disease,Department ofPhysical Medicine and Rheumatology, King's College Hospital,

and the Department of Neuropathology, Institute of Neurology, London

SUMMARY In a prospective study of 70 unselected patients with chronic liver disease, clinical signsof a peripheral neuropathy were observed in 13 patients. Abnormal nerve conduction was demon-strated in nine of these and in one further patient who had no abnormal neurological signs. Theoccurrence of a neuropathy (in patients with cryptogenic cirrhosis, haemochromatosis, active chronichepatitis as well as in alcoholic cirrhosis) could not be related to liver function, although it wasassociated with higher IgA and IgM values. Clinical diabetes was present in six of the 14 patientswith neuropathy but there was no relation in the non-diabetic patients between neuropathy andminor impairment of carbohydrate tolerance. Those with neuropathy had a significantly higherincidence of oesophageal varices and there was also a relationship to a history of previous encephalo-pathy. Sural nerve biopsy was carried out on 14 patients, eight of whom had clinical or electro-diagnostic evidence of neuropathy. Single nerve fibres were examined by teasing and in all nerves

histological evidence was found of an indolent process which had damaged whole Schwann cellsand which resulted in demyelination and remyelination. Diabetic angiopathy was not seen and axonaldegeneration, which was never severe, was found in all disease groups equally.

The occurrence of a demyelinating peripheralneuropathy in liver disease was first described byDayan and Williams (1967). Six of their 10patients had clinical signs of nerve damageincluding absent reflexes and impairment ofvibration sense or touch, although a biopsy of thesural nerve showed active segmental demyelina-tion in all. This type of Schwann cell diseasealso occurs in the neuropathy of diabetes andother conditions (Thomas and Lascelles, 1966;Gilliatt, 1969; Dayan, Gardner-Thorpe, Down,and Gleadle, 1970), but is distinct from thehistological changes of axonal degenerationusually seen in alcoholism (Victor, 1965). Thecause of the segmental demyelination observedin the patients with liver disease was not estab-lished and the present study was undertaken todetermine the frequency of peripheral nervedamage in a larger series of patients who wereinvestigated by electrodiagnostic techniques aswell as clinically and by sural nerve biopsy. Apossible relationship to the biochemical changes,including disturbances of carbohydrate metabol-

ism, which accompany liver disease was alsoexamined.

22

METHODS

Seventy patients (36 males, 34 females) with histo-logically proven cirrhosis or active chronic hepatitiswere questioned for symptoms indicative of peri-pheral neuropathy-namely, distal sensory loss,paraesthesiae, clumsiness, or weakness. Examinationof the peripheral nervous system included theappreciation of light touch, pinprick, and two-pointdiscrimination. Alteration in vibration sense wassought by timing the duration of perception of atuning fork (frequency 128/sec) placed on theolecranon, second metacarpophalangeal joint, tibialtubercles, and medial malleoli. Isolated loss orreduction of ankle jerks in patients more than 60years old was not considered abnormal.The laboratory investigations included a blood

count and liver function tests, and in 48 patients theserum immunoglobulins. In the first 23 patientsserum B12 and folate levels were measured and thebone marrow examined. An oral glucose tolerance

Protected by copyright.

on 5 August 2018 by guest.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.35.1.22 on 1 February 1972. D

ownloaded from

http://jnnp.bmj.com/

Peripheral neuropathy in chronic liver disease

test (50 g load) was also carried out except in thosepatients known to be diabetic.

ELECTRODIAGNOSTIC TECHNIQUES Nerve conductionvelocities were measured in both lateral poplitealnerves stimulating at the head of the fibula and atthe ankle. Velocities were also obtained for theupper and lower segments of left median and ulnarnerves by stimulating at the axilla, elbow, andwrist. Bipolar electrodes of 8 mm diameter carryinga square electrical pulse of 01 msec duration wereused for stimulation, and for recording surfaceelectrodes were placed over the extensor digitorumbrevis muscle in the foot, and the thenar and hypo-thenar muscles in the hand. Electromyograms wereobtained from a concentric needle electrode insertedinto the first dorsal interosseous muscle in each hand,and in the tibialis anterior and extensor digitorumbrevis muscle in each foot. The temperature of theexamination room was more than 65°F (18-3°C).The electrodiagnostic investigations were carried outwithout knowledge of the clinical findings.

Peripheral neuropathy was considered to bepresent if there was significant slowing of motornerve conduction in one nerve (30 m/sec in thelateral popliteal nerves, and 40 in/sec in the medianor ulnar nerves). Spontaneous fibrillation and posi-tive potentials in the electromyogram with themuscle at rest, and a reduced interference pattern onmaximal voluntary effort were regarded as evidenceof denervation but were not considered diagnosticof a neuropathy unless slowing of motor nerve con-duction was also present.

RESULTS

Clinical signs of neuropathy were found in 13of the 70 patients, although only two complainedof paraesthesiae and weakness. The most com-mon abnormalities were impairment of vibrationsense in the legs-present in 13 patients-andimpairment of superficial sensation in a 'gloveand stocking' distribution which was found inseven patients. The ankle reflexes were absenton both sides or obtainable only with reinforce-ment in seven patients, in four of whom therewas also impairment or absence of knee jerks.Motor weakness of the limbs (M.R.C. grade 4)was found in five patients.Nine of these patients with clinical neuropathy

also showed electrodiagnostic evidence of neuro-pathy. In addition, there was one other patientwith significant changes in motor nerve con-duction but without clinical evidence of neuro-pathy, making a total of 14 patients in the serieswith clinical and/or electrodiagnostic evidenceof a peripheral neuropathy. The relevant clinical

data for these 14 patients are given in Table 1. Innone was there clinical evidence of severe arterialdisease.Neuropathy was found most frequently in

alcoholic cirrhosis and haemochromatosis (Table2) but the numbers in some of the aetiologicalgroups were relatively small. None of thepatients with haemochromatosis had had ahistory of significant alcohol intake. The patientwith the most severe neuropathy had activechronic hepatitis, and is described in more detaillater.A comparison of those with and without

neuropathy revealed no significant differences inthe liver function tests except for changes inserum immunoglobulins (Table 3). IgA and IgMwere considerably higher in those with evidenceof neuropathy and the elevation of IgM (to morethan twice the level in the patients without neuro-pathy) reached statistical significance (P <0 05).These alterations in immunoglobulin levels wereindependent of the type of cirrhosis, and thepatients with neuropathy when considered to-gether had higher mean IgA and IgM values thanthose found in any individual disease group.

Seven (50%0) of the 14 patients with neuro-pathy showed either oesophageal varices onbarium swallow examination or had had aportacaval shunt, a significantly higher incidencethan in the patients with normal conduction,27% of whom had varices (P < 0 05). There wasalso a similar correlation with a past history ofhepatic encephalopathy, seven (50%0) of thepatients with neuropathy having had previousepisodes of encephalopathy as compared withfive (9%0) of the 56 patients without neuropathy(P<0 001). No patient showed features ofhepatic encephalopathy at the time of the study.

RELATION TO IMPAIRED CARBOHYDRATE TOLER-ANCE Of the 14 patients with a peripheralneuropathy, six were known to have diabetes.The severity of the neuropathy bore no relation-ship to the duration of diabetes. In four patientsdiabetes had been diagnosed more than 10 yearspreviously, whereas in the other two patientswith an equally severe neuropathy diabetes hadbeen present for only six months. Both of thesepatients had also had episodes of portosystemicencephalopathy, which may have contributed totheir neuropathy. Of the 56 patients withoutneuropathy four had diabetes, a significantlylower incidence (P <00 1).

Sixteen of the patients who did not have

23P

rotected by copyright. on 5 A

ugust 2018 by guest.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.35.1.22 on 1 F

ebruary 1972. Dow

nloaded from


R. P. Knill-Jones, C. J. Goodwill, A. D. Dayan, and Roger Williams

04o0RiV- +40

~..

R .t .0

.F

_.

0 4)

'4 .~4) 0

) 4)

00

-~

0204.

00t

-0 40-0 '4

0

+ + ++ + ± +±.

+ + + ± + + + ++ + + + + + + +

+ + ±

++ + ++++

+ + +

+++ +

+ ±+ +

4)4 0 00 04 - 0, 00 en r- Nen - *s ") en en *o N *s *-1 e-, ") c

N 'I 0 - 00 c t - 0 .0 . 40-'t e t e N IS v e t

t r "t It ootr x0 't - 00 00I N en T t t T - 4.t en

z 0 0z 0 z0<-~0 z z~z z

z 0 <z 0 z z< ZZ z z z

z < z 0

4 >,4)-.~N~ 4).0E )00044 0>' 0- i5 c

.2N

4c

.04~ .0 E e N e4D4 - 440o D -*a: a> @

+ 00 > O0 + 0 + C, + + 'D + 040 Q4

+ + ,0 +

+ + + +* + + + + + +

+ +** J tJ ++ +

.. > + + + +

44 O40 .0 -_

cr- O0 - _-

'IN t - .V. 4)4 ,) .0 t 4)4

z z z z z z

z z z S z z

.2.0 00I4c-._ _

0 *n N Ns)4 4t *n 0

0 0 0, 0 + 0

0 0. 0 "0 o o o. 0 - 0o o - 0;'o -''4 -

to0 u ; tw 0 ;c0 8=0 E - °=0 wo u° °E 0 ° E Eoto 0 u' 0 0 u'=YO .0 O.0000 0O Oxox0 2=0, 2 CO CZS..E0. -0 O.O0 4)4C

E- 0.-

n N 4. 0 '0 4 00n 04 0 N 4.Oun 'IC ) V) It 'IC It 'I 't 'C 't 'C It

N m It tn c 1- oo C> O - N s I'd,

C-

_ t 00 N

:z

- ;z 0Id,O

z _ W_ 'I 'I_ in

24

b4.E

.0-0

.0

14

40

40

g0t

0ri

30U

H

uzH:0

H4

x0-

LL

z

40

3>aH;

z

z0:

H

H

40

40

0

01

0

.0

0

.0

.0

11

Ea0

+

++

00

. _

0

D

0l

0

0

.0

C

0

._

0

04.0o0

0L.

c40'

;~

q

t40.

Cz0



http://jnnp.bmj.com

/J N

eurol Neurosurg P


ownloaded from



TABLE 2INCIDENCE OF NEUROPATHY IN THE DIFFERENT GROUPS

Diagnosis Patients (tno.) NeuropathY

Active chronic hepatitis 26 2Alcoholic cirrhosis 13 6Cryptogenic cirrhosis 12 3*Primary biliary cirrhosis 12 0Haemochromatosis 7 3

Total 70 14

* One of these patients had electrodiagnostic evidence of neuropathyonly.

TABLE 3COMPARISON OF CLINtCAL FEATURES, BIOCHEMICAL ANDHAEMATOLOGICAL FINDINGS IN PATIENTS WITH (14) ANDWITHOUT (56) EVIDENCE OF A PERIPHERAL NEUROPATHY

Normial NeuropathY

Age:Mean (yr) 51 ± 2-1 55 ± 2-5Range (yr) 14-79 45-67

Duration of disease (months) 27 + 3 28 ± 8Total bilirubin (mg/100 ml.) 3-1±0-6 1.2±02Aspartate aminotransferase (u./100 ml.) 77 ± 12-2 54 + 7-7Alkaline phosphatase (K.A. u./100 ml.) 26 ± 3-3 15 ±2 5Albumin (g/100 ml.) 2-9 ± 0-1 2-8 ± 0-1Total globulin (g/100 ml.) 46±+0-2 4-1 ±0-4Immunoglobulins (mg/100 ml.)IgG 2033± 148 1797 +188IgA 470± 466 785 ± 308IgM 223 ± 274 539 ± 312

B12 (Ils/ml-) 568 ± 54 627±74Folate (mtA/ml.) 5-3 ±07 5 2± 0-6

Carboh7ydrate tolerance (diabeticpatients excluded)

Blood sugar (mg/100 ml.)fasting 75± 31 75 ± 4 4i hr 125±4-5 122+10-51 hr 140±6-1 164+1601Ihr 123±6-3 137+1202hr 104±6-5 97±90

Mean values are given ± standard error of the mean.

clinical diabetes were found to have an impairedglucose tolerance test (defined as a two-hourblood sugar level higher than 120 mg/100 ml.).In each case the fasting level was below 100 mg/100 ml. One had a neuropathy and 15 did not,giving a similar incidence in both groups ofpatients. Comparison of the patients with araised two-hour blood sugar level and those witha normal level revealed no significant differencesin nerve conduction velocities, latencies, orvibration perception times (Table 4). Thus, if thepatients with clinical diabetes are excluded, thereis no relationship in the remainder betweenneuropathy and minor impairment of glucosetolerance.

SURAL NERVE BIOPSY This was obtained underlocal anaesthesia in 12 patients and at necropsy

TABLE 4COMPARISON OF MEAN NERVE CONDUCTION VELOCITIES,

MOTOR LATENCIES, AND DURATION OF VIBRATION PERCEP-

TION IN 44 PATIENTS WITH NORMAL GLUCOSE TOLERANCE,16 WITH MINOR ABNORMALITIES OF GLUCOSE TOLERANCE

AND 10 WITH DEFINITE DIABETES

Blood sugar at 2 hr Diabetes

Nornmal Abnornmal

NerLe Segnzent Nerve conduction velocity (m/sec)

Median Upper 62±1.5 55+1-4 51±49Lower 52±09 52 ± 2-0 49± 18

Ulnar Upper 58± 18 60± 15 50±2-2Lower 55±1.5 53±1-8 47±2-6

Lateral Left 44± 1-3 45±2 7 40±2-8popliteal Right 43± 1-0 44±2 8 39±2-1

Ner-e Point of Mean latencies (ni/sec.)stimulation

Median Wrist 3-7 1-2 3-3 + 2 2 3-8 ±26Ulnar Wrist 3-2±1*1 3 3±2-1 3-3±2 6Left lateral Ankle 5-6±2-7 5-3 333 6-2±5-3

poplitealRight lateral Ankle 5-6±2-2 5-33±8 6-1 ±43

Arerage of left and rightvibration duration (sec.)

Elbow 12±09 11±1-5 11±23MP joint 16±1-1 15±1-8 14±1-3Knee 9±08 9±1 4 5±1-9Ankle 10±09 9±1 1 3±2-0

Mean ± I SE of mean.

in two patients. Biopsies were obtained fromseven of the 13 patients with clinical evidence ofneuropathy and from one patient with abnormalconduction velocities only (Table 1). The re-

maining six biopsies were from patients withoutdefinite evidence of neuropathy.

All 14 biopsies showed some histologicalabnormalities, although these varied consider-ably in severity from case to case. No definitelesions were found in the nerves in standardsections, although there were apparent differencesin the amount of endoneurial fibrous tissue andin the relative proportions of the larger andsmaller diameter myelinated nerve fibres. Smallblood vessels in the perineural spaces appearednormal and did not show hyalinization of theirwalls or thickening or reduplication of the base-ment membranes. Only a very occasional de-generating fibre was seen in frozen sectionsstained to show myelin sheaths; this was no

more frequent in the patients with alcoholiccirrhosis than in the other diagnostic groups.

Isolated nerve fibres separated by freehand'teasing' showed two main types of damage.The commonest was the presence along fibres ofpartly or completely demyelinated internodes-

25P



j.com/

J Neurol N


ebruary 1972. Dow

nloaded from



.X.- - -.- .....

'I-

FIG. 1 a. Consecutive lengths of a teased nerve fibreshowing many short, thin, re-myelinated internodesformed after extensive segmental demyelination (fromcase 17, Table 1).

FIG. lb. Similar to Ja. In addition, there are somedenuded areas due to the presence of active Schwanncell-myelin damage (from case 3, Table 1).

o.. .:,,- r. aN __0 *_ k

FIG. 1 C. Consecutive lengthsof a teased nerve fibre showinggranular debris due to com-plete, Wallerian-type de-generation (from case 18,Table 1).

26

_



http://jnnp.bmj.com

/J N

eurol Neurosurg P


ownloaded from



'segmental demyelination'-together with theresults of healing of this lesion-chains of uni-formly too-short segments interspersed betweeninternodes of normal length (Fig. 1). Such healed(or healing) lesions were more frequent thanactively breaking down myelin, and the repara-tive processes appeared to be following the nor-

mal pattern.No evidence was seen of excessive prolifera-

tion of Schwann cells forming 'onion bulbs', or

of excessive accumulation of myelin breakdownproducts. The other type of damage, 'Wallerian-like' or 'total' nerve fibre degeneration, in whichboth axis cylinders and their myelin sheaths aredestroyed, was seen less often (10 biopsies),particularly in the acute stage of a long chain ofosmiophilic globules, or as a Schwann cell tubeforming a band of Briingner. Healing of thisprocess, which is necessarily slow because it re-

quires regrowth of an axonal sprout, eventuallyproduces fibres which bear stoutly myelinatedinternodes uniformly too short for a normal fibreof that diameter. The latter type of abnormalnerve fibre was relatively more frequent in thepresent biopsies than acutely degeneratingforms. The overall impression was of an indolenttype of pathological process, sometimes quiteextensive and severe (more than 500 of teasedfibres were abnormal in some cases), which hadproduced greater damage to the Schwann cell-myelin complex than to the axis cylinders.Restricted paranodal lesions of the myelinsheath were very uncommon, suggesting thatwhatever had damaged the nerves had done sowith such an intensity, or for such a duration,that entire Schwann cells had degenerated,rather than just their most distal and labile tips.The data for the internodal-length relation-

ship in four patients has been plotted for eachfibre according to the convention of Fullerton,

Gilliatt, Lascelles, and Morgan-Hughes (1965),in which individual segments on an isolatedfibre are represented by points on a vertical line(Fig. 2). Comparison with normal values (Thomasand Lascelles, 1966) shows how very extensivethe damage was in some patients, and how muchmore common were features of Schwann cell-myelin lesions than complete nerve fibredegeneration.No relationship could be found between the

histological changes in the nerves and thepresence or absence of diabetes or oesophagealvarices (Table 1).

SERIAL OBSERVATIONS AND RESPONSETO TREATMENT IN TWO PATIENTS

CASE 1 (Table 1) This male patient had beennoted to have splenomegaly at the age of 9 years

during an attack of jaundice. However, he re-mained well until the age of 52 when he presentedwith ankle oedema and mild portosystemicencephalopathy. Radiological examinationshowed an extensive collateral circulation(Blendis, Laws, Williams, and Thompson, 1968).The glucose tolerance test was normal. Therewere no signs of clinical neuropathy but the con-duction velocities were reduced in both arms andlegs and there was electromyographic evidenceof denervation. Sural nerve biopsy showed activeand healed demyelination with some loss ofnerve fibres. The patient was treated with neo-mycin and a restricted protein diet with return ofconduction velocities to normal over an 18month period (Table 5).

CASE 2 (Table 1) A sales manager aged 62 yearspresented with an eight-month history of un-

steadiness of gait and numbness in the legs andfingers. He denied excessive intake of alcohol but

TABLE 5NERVE CONDUCTION VELOCITIES AND ELECTROMYOGRAM CHANGES IN TWO PATIENTS DESCRIBED IN DETAIL IN TEXT

Conduction velocities (m/sec) Electromyogram abnormality

Left median nerve Left ulnar nerve Lat. popliteal nerve At rest Volitional

Upper Lower Upper Lower Left Right

Case 1Jan. 1968 43 36 52 44 42 35 Yes YesJune 1968 58 50 57 47 57 40 No YesJune 1969 64 59 57 63 40 49 No No

Case 2Feb. 1968 24 31 20 23 16 16 Yes YesJuly 1968 33 40 30 30 24 20 Yes YesAug. 1969 44 35 36 30 24 19 No No

27P



j.com/

J Neurol N


ebruary 1972. Dow

nloaded from



4~~~~~~~~~~~~~~~~r

.1

.siO

.'D,0 i - I

FIG. 2. Graphical analysis of sural nerve from cases 1, 2, 4, and 5 (Table 1). Each vertical line representsone fibre; the wide scatter of internodal lengths is due to segmental demyelination and remyelination. Shadedarea encloses mean and 95%0 confidence limits (Thomas and Lascelles, 1966).

gave a history suggestive of infective hepatitiswith apparently complete recovery at the age of38 years. Liver function tests and liver biopsyappearances were consistent with a diagnosis ofactive chronic hepatitis. On examination therewere absent leg reflexes, marked wasting, andgreatly impaired vibration and position sense inboth arms and legs, with marked impairment ofsuperficial sensation below the knees. Oesopha-

geal varices were not detected; serum B12 andfolate levels as well as a pyruvate tolerance curvewere all normal.

Sural nerve biopsy showed a severe degree ofactive demyelination with some healed lesionsand also some evidence of fibre loss. Conductionvelocities and the electromyogram were grosslyabnormal (Table 5). The patient's symptoms ofneuropathy improved with azathioprine but,

28

i



http://jnnp.bmj.com

/J N

eurol Neurosurg P


ownloaded from



because of nausea, treatment was later changedto prednisone (20 mg/day). Conduction veloci-ties improved further and after one year of treat-ment the knee reflex had returned and after afurther six months, the ankle reflex also. Liverfunction tests improved during this period butthe glucose tolerance tests became increasinglyabnormal, probably as a result of prednisonetherapy.

DISCUSSION

The present study confirms the earlier findings ofDayan and Williams (1967) in showing the fre-quent presence of segmental demyelination inpatients with liver disease-indeed, it waspresent in every sural nerve biopsy examined.Clinical signs of a neuropathy were less fre-quent-19%0 of the patients in the present series.Ortiz V'asquez, Olmo, Muro, Villamol, andBarreiro, 1967 have also reported clinicallyevident neuropathy in hepatic failure but manyof their patients were alcoholics and no informa-tion is given as to the presence of diabetes.Malkova (1967) found 11 patients with aperipheral neuropathy among 242 patients withliver or gall-bladder disease in Moscow. Nopatient had a significant alcohol intake and allthose with a neuropathy had spider naevi andpurpura; seven had hypergammaglobulinaemia.The only case reported in detail had a painfulperipheral neuropathy in one arm and later inone leg with an isolated blood sugar of 122 mg/100 ml. It is possible that this was diabeticmononeuritis, a type of neuropathy we have notyet seen in liver disease.The occurrence of peripheral neuropathy in

the present group of patients could be onlypartly accounted for by known causes such asalcoholism and diabetes. Furthermore, thehistological abnormalities in the patients witheither diabetes or alcoholism were similar tothose seen in the patients in whom these factorscould not be incriminated. Neither the small-vessel disease frequently seen in the nerves indiabetes nor the axonal degeneration of alcohol-ism was prominent. Indeed, the similarity of thehistological picture in all 14 patients examinedsuggests that another factor common to allpatients was the main cause of the neuropathy.The importance of collateral shunting wasclearly shown and even in the cirrhotic patientswith diabetes and in the alcoholics this may bethe more important factor. The two patients

with a severe neuropathy and collateral shuntingwho had only recently developed diabetes are agood illustration of this. Collateral shuntingallows toxic factors, normally produced in thecolon and detoxicated by the liver, access toboth central and peripheral nervous tissue(Victor, Adams, and Cole, 1965; Cavanagh andKyu, 1969). Not surprisingly this correlates withperiods of encephalopathy and this aetiologicalrelationship is further supported by the improve-ment in conduction velocities seen in case 1during protein restriction.

In one series of cirrhotic patients from GreatBritain, diabetes was present in 32% and ab-normal glucose tolerance was found in a further25% of patients (Megyesi, Samols, and Marks,1967). The impairment of glucose tolerance isassociated with relative hyperinsulinism but theexact cause is unknown. Although in ourpatients there was a higher incidence of neuro-pathy in those with clinical diabetes than inthose without, there was no correlation betweenneuropathy and minor abnormalities of glucosetolerance. This finding is in agreement with therecent study of Seneviratne and Peiris (1970)who showed that electrodiagnostic evidence ofneuropathy was common in a group of patientswith chronic liver disease selected because of anormal glucose tolerance test.Advancing age is itself associated with clinical

(Mayne, 1965) and pathological evidence (Las-celles and Thomas, 1966) of damage to peri-pheral nerves for reasons which are not known.In the present series 13 of the 18 patients withneuropathy were less than 60 years old, the agebeyond which 'physiological' nerve damagebecomes apparent clinically. Furthermore, theseverity of the clinical signs and the extent of thenerve lesions in some of the older patients weregreater than could be due to ageing alone(Arnold and Harriman, 1970).There was no relationship between the occur-

rence of neuropathy and the duration of liverdisease, or the changes in liver function tests.The high IgA and IgM levels are of interest butthe association may simply indicate that neuro-pathy is related to an 'active' stage of cirrhosisas illustrated by case 2. In this patient theneuropathy improved on corticosteroid treat-ment simultaneously with improvement in thebiochemical changes of active chronic hepatitis.Other organs including skin, lungs, and colonare not infrequently involved in this condition,such multi-system involvement implying an

29P



j.com/

J Neurol N


ebruary 1972. Dow

nloaded from



'abnormal immune system' in this patient. Thefindings in case 2 would suggest that theperipheral nerves can be added to the list oforgans sometimes involved in this disease.Alternatively, the immunoglobulin abnormalitiesmay be more directly related to the cause of thenerve damage, for neuropathy is found inpatients with various types of paraproteinaemias(Logothetis, Kennedy, Ellington, and Williams,1968) and is common in patients with myeloma(Morley and Schwieger, 1968). Unfortunately,the histological appearances themselves give noclue to the aetiology of the neuropathy in ourpatients, although they do suggest that the finalcommon path is Schwann cell damage.

REFERENCES

Arnold, N., and Harriman, D. G. F. (1970). The incidence ofabnormality in control human peripheral nerves studied bysingle axon dissection. Journal of Neurology, Neurosurgeryand Psychiatry, 33, 55-61.

Blendis, L. M., Laws, J. W., Williams, R., and Thomson,W. B. (1968). Calcified collateral veins and gross dilatationof the azygos vein in cirrhosis. British Journal ofRadiology,41, 909-912.

Cavanagh, J. B., and Kyu, M. H. (1969). Colchicine-likeeffect on astrocytes after portacaval shunt in rats. Lancet, 2,620-622.

Dayan, A. D., and Williams, R. (1967). Demyelinatingperipheral neuropathy and liver disease. Lancet, 2, 133-134.

Dayan, A. D., Gardner-Thorpe, C., Down, P. F., andGleadle, R. I. (1970). Peripheral neuropathy in uremia.Pathological studies on peripheral nerves from 6 patients.Neurology (Minneap.), 20, 649-658.

Fullerton, P. M., Gilliatt, R. W., Lascelles, R. G., andMorgan-Hughes, J. A. (1965). The relation between fibrediameter and internodal length in chronic neuropathy.Journal of Physiology, 178, 26-28P.

Gilliatt, R. W. (1969). Experimental peripheral neuropathy.In Scientific Basis of Medicine Annual Reviews 1969, pp.202-219. British Postgraduate Medical Federation.Athlone Press: London.

Lascelles, R. G., and Thomas, P. K. (1966). Changes due toage in internodal length in the sural nerve in man. Journalof Neurology, Neurosurgery and Psychiatry, 29, 40-44.

Logothetis, J., Kennedy, W. R., Ellington, A., and Williams,R. C. (1968). Cryoglobulinemic neuropathy. Incidenceand clinical characteristics. Archives of Neurology, 19,389-397.

Malkova, E. V. (1967). Syndrome of multiple neuritis indiseases of the liver. Zhurnal Nevropatologii i Psikhiat riimeni S.S. Korsakova, 67, 1189-1193.

Mayne, N. (1965). Neuropathy in the diabetic and non-diabetic populations. Lancet, 2, 1313-1316.

Megyesi, C., Samols, E., and Marks, V. (1967). Glucosetolerance and diabetes in chronic liver disease. Lancet, 2,1051-1055.

Morley, J. B., and Schwieger, A. C. (1968). Aspects of chronicpolyneuropathy associated with myeloma. Proceedings ofthe Australian Association of Neurology, 5, 403-404.

Ortiz-Vasquez, J., Olmo, A., Muro, J., Villamor, J., andBarreiro, P. (1967). Neuropatia hepatica. Revista clinicaespaniola anida, 104, 60-66.

Seneviratne, K. N., and Peiris, 0. A. (1970). Peripheralnerve function in chronic liver disease. Journal ofNeurology,Neurosurgery and Psychiatry, 33, 609-614.

Thomas, P. K., and Lascelles, R. G. (1966). The pathologyof diabetic neuropathy. Quarterly Journal of Medicine, 35,489-509.

Victor, M. (1965). In The Remote Effects of Cancer on theNervous System. Edited by Lord Brain and F. H. Norris.Grune and Stratton: New York.

Victor, M., Adams, R. D., and Cole, M. (1965). The acquired(non-Wilsonian) type of chronic hepatocerebral degenera-tion. Medicine (Balt.), 44, 345-396.

30P



j.com/

J Neurol N


ebruary 1972. Dow

nloaded from


peripheral disease - journal of neurology, …jnnp.bmj.com/content/jnnp/35/1/22.full.pdf ·...

Documents