871

Section of Neurology REFERENCES Bertrams J, Hoher P G & Kuwert E (1973) Lancet i, 1287 Bradley B A, Edwards J M & Franks D (1973) Tissue Antigens 3, 340 Brewerton D A, Nicholls A, Caffrey M, Walters D & James D C 0 (1973) Lancet ii, 994 Cazzullo C L & Smeraldi E (1972) Lancet ii, 430 Compston D A S, Batchelor J R & McDonald W I (1976) Lancet ii, 1261 DuPont B, Jersild C, Hansen G S, Staub-Nielsen L, Thomsen M & Svejgaard A (1973) Transplantation Proceedings 5, 1543 Halim K & Festenstein H (1976) Lancet ii, 1255 Jersild C, Fog T, Hansen G, Thomsen M, Svejgaard A & Du Pont B (1973) Lancet ii, 1221 Jersild C, Svejgaard A & Fog T (1972) Lancet i, 1240 Jorgenson F, Lamm L U & Kissmeyer-Nielson F (1973) Tissue Antigens 3, 323 Kissmeyer-Nielsen F ed (1975) Histocompatibility Testing. Munksgaard, Copenhagen; p 5 Lamm L U, Friederich U, Peterson G B, Jorgenson F, Nielson J, Therkelson A J & Kissmeyer-Nielsen F (1974) Human Heredity 24, 273 McDevitt H 0 & Bodmer W F (1974) Lancet i, 1269 Moller G ed (1976) Transplantation Reviews. Munksgaard. Copenhagen; p 32 Naito S, Namerow N, Mickey M & Terasaki P I (1972) Tissue Antigens 2, 1 Sachs D H & Cone J L (1973) Journal of Experimental Medicine 138, 1289 Schlosstein L, Terasaki P 1, Bluestone R & Pearson C M (1973) New England Journal of Medicine 288, 704 Singal D P, Mickey M R & Terasaki P I (1969) Transplantation 7, 246 Terasaki P 1, Park M S, Opelz G & Ting A (1976) Science 193, 1245 Yunis E J & Amos B D (1971) Proceedings of the National AcademY of Sciences USA 68, 3031 Winchester R J, Ebess G, Fu S M, Espinosa L, Zabrieskie J & Kunkel H G (1975) Lancet ii, 814

immunosuppression one must first scrutinize the validity of the evidence for a major role of the immune system in this disease. If someone unbiased by any particular school of thinking and methodological expertise were to work his way through the vast amount of available information about the possible cause of MS, he might well end up with the picture of a multifactorial ,etiology as shown in Fig 1. There is good evidence for the involvement of metabolic factors, modified by diet (Mertin & Meade 1977); for a viral infection (Johnson 1975), presumably taking place at a certain age; and for immune reactions to cause the symptoms of the disease or perpetuate it once central nervous system (CNS) tissue damage has been brought about by other factors (Paterson 1973). If the assumption of a multifactorial cause of MS were to prove correct, then subsequently any curative regimen would have to include a combination of different therapeutic approaches. This may perhaps be one of the reasons why all therapeutic trials, aiming usually for an interference with only one of these possible factors, have been so unsatisfactory up to date. Immune mechanisms in MS In recent years the notion of immune deviation has been discussed increasingly in MS research. A congenital defect or a viral infection may impair the function of certain cells of the immune system, the thymus-derived T cells (which are the agents of

the cell-mediated immune response), and thus make possible a continued virus infection of the CNS. The viruses thought most likely to be responsible are those of the paramyxo group; for AdWao. in hwtlM of Faott Ack

,bon, error of

Motabolosm

Dr J Mertin (Clinical Research Centre, Northwick Park Hospital, Harrow, Middlesex HA] 3UJ)

MUltiple

A

The Evidence Justifying Immunosuppression Therapy in Multiple Sclerosis It is a rather difficult task to elaborate on the existing evidence justifying immunosuppressive treatment in patients with multiple sclerosis (MS). The main obstacle to the removal of this difficulty lies in the still unresolved question of whether immune mechanisms are decisively involved in the natural history of MS and, if so, in what way and to what degree. Thus, in any attempt to justify

Fig 1 The multifactorial a?tiology ofmultiple sclerosis. Shaded areas indicate involvement ofgenetically determined or influenced membrane functions

872

Proc. roy. Soc. Med. Volume 70 December 1977

PRIMED B-LYMPHOCYTE + CNS ANTIGEN

TARGET CELL DESTROYEO BY:

Type II reactions are mediated by cytotoxic antibodies in combination with other factors; for Anti-glial facto,t IgG ?) example, the complement system (Fig 2). Humoral substances bringing about demyelination in CNS Anpiodydeendentpcytotonesty

ANTIGEN BEARING TARGET CELL

P10gma cells d y duction mAnto

(i

--.O.

J

~~~~~~~+Dpsornn-oc. +

Comnlement.[

Opsonic adherence to phagocytic cells

lennes adherence to phagocic

cells

Lysis

Fig 2 Cytotoxic hypersensitivity (type I1l)

tissue culture

systems,

for example Bornstein's

myelinotoxic factor (Bomstein 1973), may well be such cytotoxic antibodies. The presence of lymphocytic infiltrates in the CNS tissue in MS, similar to those found after induction of an autoimmune encephalomyelitis in experimental animals, points towards a crucial role of type IV reactions, the cellmediated immune mechanisms (Fig 3).

example, measles virus. This has been shown to be true of subacute sclerosing panencephalitis, and parallels were readily drawn to MS. However, therapeutic attempts to restore an impaired lym- Immunosuppression phocyte function with the help of a so-called The rationale for immunosuppressive therapy is to 'transfer factor' have not, up to now, shown any suppress the recurring attacks of the immune decisive clinical improvement in the MS patients system against the CNS- tissue and possibly to treated (DuPont 1975). A recent observation that break the vicious circle of recurrent antigen release MS patients are able to mount a normal immune which boosts the immune response perpetuating response to measles infection (Wroblewska et al. the disease. Both immune mechanisms described 1976) weakens the argument for crucial involve- can be influenced by interference with the activity ment of measles virus. of the cell populations involved. This interference The older and more established theory linking can be achieved by blocking specific functions of MS with the immune system envisages autoim- the cells; for example, lymphocyte-antigen intermune mechanisms bringing about the primary action by high doses of nonimmunogenic antigen demyelination. Many workers supporting this fragments (Arnon 1975). This very specific way to suppress immune tfieory assume that a primary damage to the CNS, for example by infection by a virus, would cause reactions works well in experimental allergic enthe sensitization of the immune system to CNS cephalomyelitis (EAE). In MS, however, the theracomponents (Wisniewski 1977). Immunological peutic use of, say, fragments of basic protein of the mechanisms evoked in this way would then, either myelin is still rather questionable due to inindependently, or repeatedly boosted by the per- sufficient information about the true nature of the sisting primary cause, perpetuate the disease. Anti- antigen or antigens involved in this disease. A more general and already widely practised genic material released during the immune attacks would be another boosting factor in this process. method of immunosuppression is to interfere to a How do immune mechanisms bring about tissue varying extent with the metabolism of the immunedamage? Corresponding to the classification given competent cells involved, and thus to suppress by Coombs & Gell (1975), four types of immune their activity and numbers. In order to decide reactions are able to damage target structures. whether or not such action should be taken, one Type II and type IV are the reactions most likely to has first to convince oneself -and others -that be involved in demyelinatory processes in MS. immune mechanisms such as those described are likely to be crucially involved in the disease. What now is the evidence for their involvement in MS? There are five major aspects supplying some supPRIMED T-LYMPHOCYTE PROLIFERATIVE RESPONSE and/or CNS ANTIGEN PRODUCTION/RELEASE OF SOLUBLE FACTORS porting evidence: (1) similarities between EAE and causing: Cytotoxicity MS (Wisniewski 1977); (2) occurrence of oligoclonal immunoglobulin and an altered ratio of MacTDphRGe CActivation MarpaeElMigration ihbto immunoglobulin kappa: lambda chains in the Lymphocyte mitosis cerebrospinal fluid of MS patients (Thompson Chemotaxis etc. +

8

CYTOTOXIC RESPONSE

Cell death

Cytoto

Ic cell

Antigen bearing target cell

Fig 3 Cell-mediated hypersensitivity (type IF)

1977); (3) lymphocyte sensitization to CNS material (Paterson 1973); (4) association of certain histocompatibility antigens with MS (Jersild et al. 1975); (5) the therapeutic effect of immunosuppression in MS patients. It was the similarities between the pathological and clinical events in acute MS attacks and those of EAE which gave the first strong support to the assumption of immune mechanisms playing a role in MS. EAE is mediated by T lymphocytes. Sup-

Section of Neurology

pressing these cells either by irradiation or by suppressive drugs or - more specifically - by neonatal thymectomy or antilymphocyte globulin, results in suppression of EAE. Recently, McFarlin et al. (1974) have described relapsing EAE and this allows one to draw an even closer parallel between this experimental disease and MS. The presence of oligoclonal IgG in the cerebrospinal fluid (CSF) of MS patients is to date the best evidence for the importance of the immune system in MS. Results of a few in vivo and many in vitro studies have indicated that sensitization of lymphocytes to CNS components in MS does occur. There are arguments about whether this sensitization is specific for MS or is simply a consequence of CNS tissue damage as brought about by a variety of other conditions. Recent observations such as those by Sheramata et al. (1974) indicate a specific lymphocyte sensitization to basic myelin protein fractions in relation to clinical attacks of MS. The association of certain histocompatibility antigens in MS has already been discussed by Dr Sachs. The last and, in the context of this paper, most relevant point is the effect immunosuppression has already been demonstrated to exert on the clinical course of MS. Besides the concept mentioned before of a blockage of specific lymphocyte functions, three major possibilities have been cited in recent years (Table 1). At the usual doses, the value of steroids and corticotropin as specific immunosuppressive agents seems doubtful. Suppressive drugs such as azathioprine are general antimetabolites and therefore do not only affect immune-competent cells. In contrast, antilymphocyte globulin (ALG) has a rather selective effect on the lymphocytes (Lance et al. 1973). The supplementation of the patient's diet with essential fatty acids (Millar et al. 1973) may be another possibility to interfere with immune mechanisms as suggested by the results of experimental research in our laboratory (Mertin & Meade 1977). At first glance at the published results of the clinical trials investigating the therapeutic effects of these substances - given either alone or in combination - the evidence to date supporting Table I Immunosuppression in MS patients Adrenocortical hormones Corticotropin Drugs affecting lymphocyte metabolism e.g. azathioprine Antilymphocyte globulin

(ALG)

References Rose et al. (1970) Cendrowski (1971) Frick (1976) Swinburn & Liversedge (1973) Brendel et al. (1972) Seland et al. (1974) Lance etal. 1975) Ring et al. (1976)

873

immunosuppression of MS patients seems rather weak. Not one of the various studies can persuade a critical reader to believe that, by the introduction of immunosuppression into the treatment of MS, the patients' chances for an eventual halt to the progression of their disease have significantly increased. However, at second glance one discovers that, although the over-all beneficial effect of immunosuppression may not have been overwhelming, in most of the relevant reports there are always a few patients who have done rather well in terms of decreased relapse rates or clinically less fierce, 'smoother' attacks. This observation may indicate the existence of certain sub-groups of patients displaying the same state of the disease but reacting differently to different treatment regimens. Support for this assumption comes from recent findings in the histocompatibility field. In various immunological and nonimmunological diseases, combinations of certain tissue types are found in different patient groups, and the presence or the lack of certain tissue alloantigens can be associated with alterations in such important components of the organism as the complement system. Genetically determined alloantigens may be closely connected functionally with membrane components of specific functions - such as receptors for antibodies, hormones, drugs and viruses - and may interfere with the ligand-receptor interaction (Svejgaard & Ryder 1976). In schizophrenia, another of the various diseases with histocompatibility implications, it has already been observed that only patients of a certain tissue type react favourably to therapy with chlorpromazine (Smeraldi & Scorza-Smeraldi 1976). It seems conceivable that in MS there may also exist subgroups determined by the presence of antigenic membrane components and these groups may react differently to certain treatment regimens. First attempts have been made only recently to correlate characteristics other than the clinical state or course of the disease with the effect of immunosuppressive treatment in MS. Seland et al. (1974) found that the therapeutic effect of a combination of ACTH and ALG was best in patients with an elevated ratio of y-globulin to total protein in the CSF. In a pilot study (Knight et al. 1975) and in the present double blind trial carried out at the Clinical Research Centre, using ALG in combination with prednisolone and azathioprine, there is some indication that there may be two groups of MS patients reacting immunologically differently to the treatment, and it will be important to find out whether this difference may perhaps be correlated to differences in tissue types and other factors such as immunoglobulin or complement concentrations. If the possibility of such correlations is considered and recent develop-

874

Proc. roy. Soc. Med. Volume 70 December 1977

ments are taken into account in the decision about the treatment of selected MS patients, then immunosuppression is thoroughly justified. The restrictions I have given imply, nevertheless, that at the present time one cannot advocate the general application of immunosuppression as a standard form of MS treatment. However, in the framework of strictly controlled and monitored clinical trials the effects of immunosuppression in MS patients should be studied further, and this may not only be of some benefit for, I hope, an increasing number of patients, but may eventually also add to our knowledge about the pathological processes underlying this disease.

Swinburn W R & Liversedge L A (1973) Journal of Neurology, Neurosurgery and Psychiatry 36, 124-126 Thompson E J (1977) British Medical Bulletin 33, 28-33 Wisniewski H M (1977) British Medical Bulletin 33, 54-59 Wroblewska Z, Gilden D, Lisak R & Koprowski H (1976) The New Enqland Journal of Medicine 295, 732

Dr R A C Hughes, Dr I Gray, Dr R Clifford-Jones and Dr M Stern (Guy's Hospital Medical School, London Bridge, London SEJ 9RT) Immune Response to Myelin Basic Protein in Multiple Sclerosis

REFERENCES Arnon R (1975) In: Multiple Sclerosis Research. Eds. A N Davison et al. HMSO, London; pp 271-283 Bornstein M B (1973) In: Progress in Neuropathology. Vol 2. Ed. H M Zimmerman. Grune and Stratton, New York; p 69 Brendel W, Seifert J & Lob G (1972) Proceedings of the Royal Society of Medicine 65, 531-535 Cendrowski W S (1971) Acta neurologica Scandinavica 47, 254-260

CoombsRRA&GellPGH, (1975) In: Clinical Aspects of Immunology. 3rd edn. Eds. P G H Gell et al. Blackwell, Oxford &c.; p 761 DuPont B (1975) In: Multiple Sclerosis Research. Eds. A N Davison et al. HMSO, London; pp 291-314 Frick E (1976) Nervenarzt 47, 424-428 Jersild C, DuPont B, Fog T, Platz P J & Svejgaard A (1975) Transplantation Reviews 22, 148-163 Johnson R T (I1975) In: Multiple Sclerosis Research. Eds. A N Davison et al. HMSO, London; pp 155-183 Knight S C, Lance E M, Abbosh J, Munro A & O'Brien J (1975) Clinical and Experimental Immunology 21, 23-31 Lance E M, Kremer M, Abbosh J, Jones V E, Knight S & Medawar P B (1975) Clinical and Experimental Immunology 21, 1-12 Lance E M, Medawar P B & Taub R N (1973) Advances in Immunology 17, 1-92 McFarlin D E, Blank S E & Kibler R F (1974) The Journal of Immunology 113, 712-715 Mertin J & Meade C J (1977) British Medical Bulletin 33, 67-71 Milar J H D, Zilkha K J, Langman M J S, Payling Wright H, Smith A D, Belin J & Thompson R H S (1973) British Medical Journal i, 765-768 Paterson P Y (1973) Journal of Chronic Diseases 26, 119-126 Ring J, Seifert J, Angstwurm H, Frick E, Mertin J, Brass B, Backmund H & Lob G (1976) Postgraduate Medical Journal 52, 123-128 Rose A S, Kuzma J W, Kurtzke J F, Namerow N S, Sibley W A & Tourtelotte W W (1970) Neurology 20, 1-59 Seland T P, McPherson T A, Grace M, Lamoureaux G & Blain J G (1974) Neurology 24, 34-40 Sheramata W, Cosgrove J B R & Eylar E H (1974) The New England Journal of Medicine 291, 14-17 Smeraldi E & Scorza-Smeraldi R (1976) Nature 260, 532-533 Svejgaard A & Ryder L P (1976) Lancet ii, 547-549

For forty years cavalry brigades of research workers have attacked the fortress of multiple sclerosis (MS), spurred on by its alleged similarity with experimental allergic encephalomyelitis (EAE). The histological brigade report that EAE in large primates and human post-rabies vaccine encephalomyelitis resemble the active lesions of multiple sclerosis. The experimental pathologists have kept the battle going with descriptions of chronic, even relapsing, forms of EAE in the guinea-pig (Snyder et al. 1975). The brigade of immunologists have successfully demonstrated that EAE is caused by a T-cell mediated immune response against myelin basic protein (molecular weight 18 000). Antibody to myelin basic protein is produced but is not necessary for disease production. Antibodies to other myelin antigens are also produced but do not induce disease, and may even have a protective effect (Hughes 1974, Hughes & Leibowitz 1975). The onus is now on the immunological cavalry to identify similar phenomena in MS. The most easily detectable marker of any immune response to basic protein would be antibody. Attempts to find such antibody by radioimmunoassay in the serum of MS patients have been uniformly unsuccessful (Lisak et al. 1968, Lennon & Mackay 1972, Schmid et al. 1974). The cerebrospinal fluid (CSF) of MS patients contains oligoclonal bands of IgG which are not present in the serum, but antibody to basic protein has not been found in these bands. Possibly the excess basic protein present in the CSF of patients with active MS (Cohen et al. 1976) might mask the presence of antibodies. A multitude of different techniques have been used to detect cell-mediated immunity to basic protein, with conflicting results (Hughes & Leibowitz 1977). The most suggestive work is that of Sheremata et al. (1976) who claim that cellmediated immunity to basic protein is indeed present at the time of or shortly before a relapse of