significance immunoglobulin - bmj

Journal ofNeurology, Neurosurgery, and Psychiatry, 1979, 42, 179-183

Significance of immunoglobulin deposition inperipheral nerve in neuropathies associated withparaproteinaemiaM. SWASH, J. PERRIN, AND M. S. SCHWARTZ

From the Departments of Clinical Immunology and Pathology, The London Hospital Medical College,and the Department of Neurology, The London Hospital, London

SUMMARY Direct and indirect immunofluorescent studies of sural nerves were carried outin two patients with paraproteinaemia and neuropathy, in four other patients with axonal ordemyelinating neuropathies, and in one normal sural nerve. IgM was demonstrated directly inthe two cases of paraproteinaemia and neuropathy, and indirectly, using the serum of one ofthese cases, in a case of axonal neuropathy and in one case of chronic Guillain-Barre syn-

drome. In the latter case, IgM deposition also occurred after exposure to normal serum. Theseresults suggest that the paraprotein itself did not directly cause neuropathy, but that immuno-globulin deposition is probably a secondary process, caused by diffusion into damaged nerves.

Peripheral neuropathy occurs in association withparaproteinaemias of various causes (McLeod andWalsh, 1975; Read et al., 1978) but the patho-genesis of these neuropathies is obscure. In thispaper we report experiments on the immunofluor-escent binding of the paraprotein from two patientswith paraproteinaemia and neuropathy to theirown sural nerves, and to the sural nerves of fiveother patients, four with demyelinating or axonalneuropathies not associated with paraprotein-aemias, and one without neuropathy or parapro-teinaemia. Immunofluorescent binding of normalserum to these nerves was also studied.

Case reports

CASE IA 66 year old man presented with weakness andtingling in his legs. A chest radiograph in May1976 revealed a small lobulated opacity in theright lower lobe but bronchoscopy and sputumcytology were both negative. The ankle jerks wereabsent, and there was a mild distal impairment ofsensation, with a slightly ataxic gait. In May 1977he was reassessed because of increasing difficultywith walking. There was slight weakness of theintrinsic muscles of both hands, with impaired two-

Address for correspondence and reprint requests: Dr M. Swash,Department of Neurology, The London Hospital, London, El IBB.Accepted 12 September 1978

point discrimination in the fingers; in the legsthere was distal weakness and sensory impairment.The ankle jerks were absent but the other reflexeswere just present. The left median motor nerve con-duction velocity was 19 m/s, and an F wave studyrevealed a proximal conduction velocity of 25 m/s.The left sural sensory nerve action potential wasabsent. The chest radiograph showed no change inthe small lesion in the right lower lobe. The ESRwas 65 mm/hr. Serum protein electrophoresisshowed a paraprotein. Immunoelectrophoresis re-vealed an IgA level of 1.7 g/l (normal 1.7-4.1 g/l),IgG 11.0 g/l (normal 6.0-12.0 g/l), and IgM9.0 g/l (normal 0.6-1.8 g/l). The paraprotein con-sisted of IgM heavy chains and kappa light chains,and the urine contained Bence-Jones protein,which consisted of kappa light chains. A mito-chondrial antibody was found in the patient'sserum. This mitochondrial antibody was found inthe IgA and IgG fractions, but not in the IgMkappa paraprotein. Marrow biopsy and a radio-graphic bone survey showed no evidence ofmultiple myelomatosis. Sural nerve biopsy was per-formed. In May 1978, repeat bronchoscopy showedpoorly differentiated squamous cell carcinoma.

CASE 2For a year this 69 year old man had noticed pro-gressive distal numbness and paraesthesiae. Duringthis time his walking had deteriorated. On exam-

179G

Protected by copyright.

on Decem

ber 2, 2021 by guest.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.42.2.179 on 1 F

ebruary 1979. Dow

nloaded from

http://jnnp.bmj.com/

M. Swash, J. Perrin, and M. S. Schwartz

ination there was a glove and stocking sensorydisturbance with distal weakness. The tendon re-flexes were absent. The ESR was 75 mm/hr and theCSF protein was 1.4 g/l. The right median motornerve conduction velocity was 30 m/s. Median andsural nerve sensory action potentials could not beobtained. Serum protein electrophoresis showedincrease in gamma globulins with an IgM levelof 4.0 g/l. This IgM was polyclonal. Sural nervebiopsy was performed.

Methods

The sural nerve biopsy samples were embedded inAraldite and prepared for light and electronmicroscopy. In addition, small pieces of thesespecimens were snap frozen in isopentane/liquidnitrogen. It was thus possible to study frozen sec-tions of peripheral nerve from the two cases ofneuropathy associated with paraproteinaemia(cases 1 and 2), and to compare their immuno-fluorescent reactions with nerve biopsy samplesfrom two cases of Guillain-Barre syndrome (cases3 and 4) and cases of idiopathic axonal neuropathy(case 5) and Dejerine-Sottas disease (case 6). Anormal sural nerve (case 7) was obtained atnecropsy from a patient who had died of myo-cardial infarction (Table 1).

IMMUNOFLUORESCENCECryostat sections 6 ,um thick, were stained by thedirect immunofluorescent technique, using sheepantihuman fluorescent-labelled conjugates to IgG,IgM, IgA, and rabbit antihuman conjugates to C3,fibrin, and albumin. In separate experiments cryo-stat sections of these nerves were first exposed tocase 1 serum or to a normal serum and stained bythe indirect immunofluorescent technique usingsheep or rabbit antihuman IgM fluorescent conju-gate, as in the direct technique. The stained sec-tions were examined by ultraviolet epi-illuminationusing a Zeiss universal research microscope.

In case 1 the IgM fraction was separated on a

Sephadex G200 chromatograph column: 7S IgMwas identified by the method of Stobo and Tom-masi (1967).

Results

NERVE PATHOLOGYIn case 1 there was a marked loss of myelinatedfibres (Fig. 1). Many large axons were devoid ofmyelin, and others had abnormally thin myelinsheaths. Remaining myelin rings were often dis-torted or reduplicated. No inflammatory infiltratesor amyloid deposits were seen. Electron micro-

Fig. 1 Case 1. Sural nerve transverse section.Toluidine blue, original magnification X560. Fewmyelin rings remain: some are thinner than normal.Unmyelinated large axons are present.

scopy showed degenerative changes in some myelinrings. There were increased amounts of endo-neurial collagen, and the perineurium was slightlythickened. Onion bulbs were not present, and un-myelinated nerve fibres were normal. These ap-pearances were consistent with a neuropathy ofpredominantly demyelinating type.

In case 2 there were similar abnormalities. Noinflammatory infiltrates or amyloid deposits wereseen.

Cases 3 and 4 showed the typical features ofchronic Guillain-Barre syndrome including seg-mental demyelination with sparse inflammatorycell infiltrates, loss of myelin rings, and onion bulbformation. In case 5 there was axonal loss withregenerating axonal clusters. Remaining fibres werenormally myelinated. In case 6 there was a severedemyelinating neuropathy with prominent onion-bulb formation. There was no amyloid. In case 7the sural nerve was normal.

IMMUNOFLUORESCENCEDirect immunofluorescenceIn cases 1 and 2 there was positive staining forIgM in the myelin sheaths of the degeneratingnerve. The staining, most clearly identified in thecross section of the nerve, formed a collar in thethickness of the myelin sheath round the central,unstained axis cylinder (Fig. 2). The IgM depositedin the myelin had the same kappa specificity as theserum paraprotein and as the urinary Bence-Jonesprotein (Table 1). In case 2, although the serumIgM was polyclonal, with a mixture of kappa andlambda light chains, there was no staining of themyelin sheaths in the sural nerve biopsy specimenwith either anti-kappa or anti-lambda conjugates.The IgM staining in this case was much weakerthan in case 1, with less deposition of the immuno-globulin, and this probably accounts for the nega-

180


on Decem


j.com/

J Neurol N


ebruary 1979. Dow

nloaded from


Significance of immunoglobulin deposition in peripheral nerve in neuropathies

Fig. 2 Case 1. Direct immunofluorescence for IgMin sural nerve, original magnification X320. Theperineurium and the whole thickness of theremaining myelin sheaths show fluorescent staining.Transverse and longitudinal sections.

tive reaction for kappa and lambda light chaincomponents. The remaining cases showed no im-munoglobulin deposition in the myelin sheaths or

elsewhere in the nerve (Table 1). No nerve showeddetectable deposition of C3, albumin, or fibrin.

Indirect immunofluorescenceTo ascertain if the IgM-kappa paraprotein in case1 had any affinity for normal or abnormal nerve,a number of the sural nerve samples were exposedin vitro to serum from case 1, and to a controlserum with a normal level of polyclonal IgM (seeTable 2). In case 4, one of the two patients withGuillain-Barre syndrome, there was positive stain-ing of both myelin and perineurium after treat-ment with both case 1 serum (Fig. 3), and normalserum (Fig. 4). Staining was more intense withcase 1 serum. In the other case of Guillain-Barresyndrome (case 3) there was no staining of thenerve by either case 1 or normal serum. In case 5,an axonal neuropathy, there was also positivestaining after treatment with case 1 serum, butnot after normal serum. Indirect staining in cases4 and 5, was patchy, and appeared to be locatedsuperficially on the myelin sheath (Figs. 2 and 3)in contrast to the denser, full thickness stainingseen by direct immunofluorescence in case 1(Fig. 1).

Table 2 Results of fluorescent staining for IgM inbiopsy treated with &ase I serum, and normal serumspecimens

Case Case I Normal Controlnumber serum serum (anti-IgM)

IgM-kappa polyclonal IgM (alone)

3 - _ _

4 + + _5 + _ _

6 - _ _

Normalnerve

Table 1 Results of direct immunofluorescent staining of sural nerve biopsies

Case Diagnosis Immunoglobulinnumber

Heavy chain Light chain C3

G M A kappa (k) lambda

I Peripheral neuropathyIgMk paraproteinaemiabronchogenic carcinoma - + - +

2 Peripheral neuropathypolyclonal IgM myeloma - +

3 ChronicGuillain-Barr6 I

4 Jsyndrome5 Severe idiopathic All negative*

axonal neuropathy6 D6jerine-Sottas disease7 Normal sural nerve

*Fluorescent stains for fibrin and albumin were negative in all these cases.

181


on Decem


j.com/

J Neurol N


ebruary 1979. Dow

nloaded from


M. Swash, J. Perrin, and M. S. Schwartz

Fig. 4 Case 4. Chronic Guilla n-Barrc syndrome.Longitudinal section, original magnification X450.Indirect immunofluorescence for IgM in sural nerveafter exposure to normal serum. There is faint stainingof myelin sheaths.

Fig. 3 Case 4. Chronic Guillain-Barre syndrome.Transverse and longitudinal sections, originalinagnification X450. Indirec t immunofluore.scence forIgM in sural nerve after exposure to case I serum.

There is fluorescence in perineurium, and on thesurface of the remaining myelin sheaths.

Discussion

IgM was detected in the myelin sheaths and peri-neurium of the sural nerve specimens by directimmunofluorescence in both the patients (cases 1

and 2) with peripheral neuropathy and parapro-teinaemia. In case 1 this IgM was the same kappa-specific monoclonal paraprotein present in theserum, and in the urine. In case 2 IgM was de-posited in insufficient amounts to allow identifi-cation of its light chain polyclonality. In previousreports monoclonal IgM-kappa has been demon-strated by the direct immunofluorescent techniquein the myelin sheath (Propp et al., 1975; Hobbset al., 1975), and in the endoneurium and peri-neurium (Chazot et al., 1974, 1976) in patients withWaldenstrom's macroglobulinaemia and peripheralneuropathy. In addition, monoclonal IgG and IgAhas been found in nerve biopsy samples of patientswith myeloma and peripheral neuropathy (Chazotet al.. 1976). However, peripheral neuropathy is an

uncommon complication of paraproteinaemia.Chazot et al. (1976) found an incidence of only9.8% in a series of 1602 patients with myeloma,Waldenstrom's macroglobulinaemia, and benignparaproteinaemia, and the significance of immuno-globulin binding to nerve in such cases is, there-fore, uncertain.The presence of immunoglobulin bound to per-

ipheral nerve might indicate a specific immuno-logical reaction against antigen in the myelinshcath, endoneurium, or perineurium. This seemsunlikely, however, since these structures differ anti-genically. Moreover, practically all autoantibodiesin sustained autoimmune responses are found inseveral immunoglobulin classes. For example, incase 1, a mitochondrial antibody was found in theIgA and IgG fractions, but not in the IgM frac-tion-that is, the autoantibody was polyclonal inrespect of heavy chain class. On the other hand,monoclonal paraproteins have recently been shownto have identifiable antibody specificities of hap-tenic type (Seligmann and Brouet, 1973; Potter,1977). Another indication of possible specific anti-body activity in tissue-deposited protein is the con-comitant demonstration of complement, but no C3was detected in either case 1 or case 2.Van Lis and Jennekens (1977) showed that

several plasma proteins, including immunoglobu-lins, could be identified in the epineurium, peri-neurium, and endoneurium of both abnormal andnormal sural nerve biopsy samples. The distribu-tion varied according to the pathology, and themolecular weight of the protein. They suggestedthat endoneurial blood vessels function as a barrierwith graded permeability. In pathological con-ditions protein leakage might be increased, andreabsorption from nerves to vessels interrupted.

182


on Decem


j.com/

J Neurol N


ebruary 1979. Dow

nloaded from


Significance of immunoglobulin deposition in peripheral nerve in neuropathie.J1

Olsson (1968) demonstrated abnormal perme-

ability to albumin in peripheral nerves, particularlyacross endoneurial blood vessels, in rats treatedwith isoniazid for two weeks, suggesting that thisneuropathy is associated with changes in the blood-nerve barrier. This mechanism might account fordeposition of IgM in the abnormal nerves of our

cases 1 and 2. Indeed, in our case 1 the neuropathywas predominantly demyelinating, and was associ-ated not only with paraproteinaemia, but with car-

cinoma of the bronchus. The neuropathy associ-ated with paraproteinaemia in myeloma andWaldenstrom's disease is typically axonal in type(McLeod and Walsh, 1975) although demyelinatingneuropathy has rarely been reported with multiplemyeloma (Mayo et al., 1965). In case 2 the EMGand pathological findings were more typical ofthe axonal neuropathy usually associated withparaproteinaemia.Our indirect immunofluorescent studies (Table

2) indicate that immunoglobulins are deposited,or develop affinity for peripheral nerve com-

ponents, as a secondary phenomenon in nerves

rendered susceptible by particular types of damage.A nerve from one of the patients (case 4) withchronic Guillain-Barre syndrome (see Luitjen andBaart de la Faille-Kuyper, 1972), and the nerve

from case 5, with an axonal neuropathy, bothshowed affinity for the monoclonal IgM-Kappa incase I serum. The other cases, and a normal nerve,

as also shown by Propp et al. (1975), showed no

such affinity. None of these nerves showed directimmunofluorescent staining (Table 1). Further, incase 4 indirect staining was present after exposure

of the nerve specimen to normal serum. Theseexperiments indicate an affinity of IgM (from case

I serum) for myelin in both demyelinating andaxonal neuropathies, suggesting that this affinityis not a specific autoimmune response. IndeedCarter (1977) found that 39 of 63 IgM paraproteinsshowed affinity for a wide range of axonal andglial antigens, but no such affinity was found inover 150 IgG or IgA paraproteins, or in patientswith increased levels of polyclonal IgM.We, therefore, suggest that there is no direct

or causal relation between neuropathy and para-

proteinaemia, but that the neuropathy is an

associated feature of the underlying disorder,similar to the neuropathy associated with malig-nant disease. Immunoglobulin (IgM) deposition isprobably a secondary process, accounted for bydiffusion into damaged nerves, and by affinity be-tween the IgM and peripheral nerve components,especially myelin sheath and perineurium.

References

Carter, P. M. (1977). Monoclonal proteins. In Im-munology in Medicine, p. 980. Edited by E. J. Hol-borow and W. G. Reeves. Academic Press: London.

Chazot, G., Berger, G., Bady, B., Dumas, R. Creysel,R., Tommasi, M., Schott, B., and Girard, P. (1974).Neuropathie peripherique au cours des dysglobu-linemies malignes: aspects immunopathologiques. LaNouvelle Presse Medicale, 3, 1355-1358.

Chazot, G., Berger, B., Carrier, H., Barboret, C.,Bady, B., Dumas, R., Creysel, R., and Schott, B.(1976). Manifestations neurologiques des gammo-pathies monoclonales. Revue Neurologique, 132,195-212.

Hobbs, J. R., Carter, P. M., Cooke, K. B., Foster, M.,and Oon C-J. (1975). IgM paraproteins. Journal ofClinical Pathology, 28, Supplement 6, 54-64.

Luitjen, J. A. F. M., and Baart de la Faille-Kuyper,E. H. (1972). The occurrence of IgM and complementfactors along myelin sheaths of peripheral nerves-an immunohistochemical study of the Guillain-Barresyndrome. Journal of the Neurological Sciences, 15,219-224.

Mayo, C. A., Daniels, A., and Barron, K. D. (1965).Polyneuropathy in the osteosclerotic form of mul-tiple myeloma. Transactions of the A merican Neuro-logical Association, 93, 240.

McLeod, J. G., and Walsh, J. C. (1975). Neuropathiesassociated with paraproteinaemias and dysprotein-aemias. In Peripheral Neuropathy, vol. 2, pp. 1012-1029. Edited by P. J. Dyck, P. K. Thomas, andE. H. Lambert. W. B. Saunders: Philadelphia.

Olsson, Y. (1968). Studies on vascular permeabilityin peripheral nerves 3 Permeability changes in vasanervorum and exudation of serum albumin in INH-induced neuropathy in the rat. A cta Neuropatho-logica (Berlin), 11, 103-112.

Potter, M. (1977). Antigen-binding myeloma proteinsof mice. In Advances in Immunology, vol. 25, pp.141-211. Edited by H. G. Kunkel and F. J. Dixon.Academic Press: London.

Propp, R. P., Means, E., Deibel, R., Sherer, G., andBarron, K. (1975). Waldenstrom's macroglobulinemiaand neuropathy: deposition of M-component on my-elin sheaths. Neurology (Minneapolis), 25, 980-988.

Read, D. J., Vanhegan, R. I., and Matthews, W. B.(1978). Peripheral neuropathy and benign IgG para-proteinaemia. Journal of Neurology, Neurosurgery,and Psychiatry, 41, 215-219.

Seligmann, M., and Brouet, J. C. (1973). Antibodyactivity of human myeloma globulins. Seminars inHematology, 10, 163-177.

Stobo, J. D., and Tommasi, T. B. (1967). A low mole-cular weight immunoglobulin antigenically relatedin 19S IgM. Journal of Clinical Investigation, 46,1329-1337.

Van Lis J. M. J., and Jennekens, F. G. I. (1977).Plasma proteins in human peripheral nerve. Journalof the Neurological Sciences, 34, 329-341.

183


on Decem


j.com/

J Neurol N


ebruary 1979. Dow

nloaded from


significance immunoglobulin - bmj

Documents