57858ournal ofNeurology, Neurosurgery, and Psychiatry 1994;57:578-581

Immunoadsorption therapy for myasthenia gravis J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.57.5.578 on 1 May 1994. Downloaded from

Noritoshi Shibuya, Takeshi Sato, Mitsuhiro Osame, Toshihiko Takegami, Shizuki Doi, Sachiko Kawanami

Abstract neurological disorders. A major disadvantage The results of a multicentre trial were of currently available plasma exchange proce- analysed to evaluate the efficacy of dures is the non-selective removal of essential immunoadsorption therapy for severe plasma components, necessitating a supply of generalised myasthenia gravis. Twenty plasma products to serve as a substitute fluid. patients with myasthenia gravis who were Plasma exchange is costly and carries serious concurrently receiving high dose pred- risks of anaphylactic reactions and viral infec- nisolone and azathioprine therapy were tions.6 9 10 It is preferable to remove pathogenic treated with an affinity-type adsorbent, substances from the circulation selectively. A using tryptophan-linked polyvinyl alco- specially designed affinity-type immunoadsor- hol gel (IM-TR), according to a standard- bent to selectively remove AChR antibodies ised treatment protocol. The 20 patients has been developed."' 12 This material is a syn- received five adsorption treatments within thetic resin consisting of tryptophan-linked a period of 10 days. In 11, pronounced polyvinyl alcohol gel (IM-TR). It adsorbs a improvement of myasthenic weakness large number of the AChR antibodies through was seen and long-term remission was hydrophobic interaction and rapidly improves maintained. The treatment was espe- myasthenic weakness.13'6 cially effective in patients with thymic The present study was carried out at six hyperplasia. Circulating acetylcholine neurological hospitals, according to a stan- receptor (AChR) antibodies were dardised treatment protocol to assess the clin- reduced by about 60% by treating one ical usefulness of IM-TR therapy in cases of plasma volume. There was no difference severe myasthenia gravis. in the rate of removal of the AChR anti- bodies between patients with thymic hyperplasia and patients with thymoma. Patients and methods Department of included in this study had severe gen- Neurology, Kawatana No serious complications occurred dur- Patients National Hospital, ing 100 procedures. It was concluded that eralised myasthenia gravis diagnosed by typi- Nagasaki, Japan the inmunoadsorption therapy with IM- cal clinical signs and raised titres of antibodies N Shibuya TR is useful in controlling symptoms in to AChR (more than 0-5 nmol/l bungarotoxin Department of patients with severe myasthenia gravis binding); patients were in myasthenic crisis http://jnnp.bmj.com/ Neurology, Juntendo University, Tokyo, who are otherwise unresponsive. with rapid deterioration within less than two Japan weeks. Criteria for exclusion were chronic T Sato (7 Neurol Neurosurg Psychiatry 1994;57:578-581) severe and stable myasthenia gravis; multiple Third Department of changes in the immunosuppressive medica- Internal Medicine, tion before treatment; and malignant (inva- Kagoshima University, Circulating antibodies to the acetylcholine sive) thymoma as determined by a are in serum in diagnosis of the thymus gland. Kagoshima, Japan receptor (AChR) detectable pathological on September 27, 2021 by guest. Protected copyright. M Osame 85%-90% of patients with generalised myas- Twenty patients, range 15-65 years (five Department of thenia gravis.'-3 Most AChR antibodies in IgG men, 20 women), have been treated with the Neurol6gy, Nagoya were acetyl- National Hospital, bind to the main immunogenic region of the a immunoadsorption. All receiving Aichi, Japan subunit of the AChR, and cause AChR degra- cholinesterase inhibitors. The dose was kept T Takegami dation secondary to cross-linking and modu- constant during the study. All patients had to Department of lation.2-5 Plasma exchange has been shown to have undergone thymectomy at least three Neurology, Sapporo induce a recovery from myasthenic months before the study. Their thymic histol- Minami National rapid Hospital, Hokkaido, weakness in association with the decline of the ogy showed thymic hyperplasia in 13, thy- Japan AChR antibodies." At a consensus develop- moma in six, and was indeterminate in one S Doi ment conference, plasma exchange was con- (table 1). The histological diagnoses were First Department of sidered effective in the management of certain obtained from the surgical pathologists' Internal Medicine, Fukuoka University, Fukuoka, Japan Table 1 Analysis of relation between clinicalfeatures and outcome ofimmunoadsorption therapy Kawanami S Removal of Correspondence to: Sex Duration Time before Thymic histology* AChR Ab Dr N Shibuya, Departnent No of ofillness thymectomy on day 10 of Neurology, Kawatana patients Age M F (y) (y) Hyperplasia Thymoma (%) National Hospital, 2005-1 Effect Shimogumigo, Kawatana, Pronounced 11 40(13) 2 9 9(5) 4(3) 9 1 58(9) Higashisonogi-gun, Poor and 9 36(17) 3 6 5(5) 3(4) 4 5 65(14) Nagasaki, 859-36 Japan. unresponsive Received 13 April 1993 p Value NS NS NS NS <0 05 NS and in revised form 7 June 1993. Values are means (SD); NS = non-significant; p values are one-tailed. Accepted 14 July 1993 *Thymic histology was unknown in one patient. Immunoadsorption therapyfor myasthenia gravis 579

reports. The thymoma were of both lympho- laboratory findings were obtained at 0, 4, 11, J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.57.5.578 on 1 May 1994. Downloaded from cytic and epithelial types. In the non-tumour 14, and 28 days according to the study proto- cases the thymus showed hyperplasia of col. The observation period was extended up lymph follicles. to 42 days if no change in the treatment had Immunosuppressive treatment with aza- occurred. The effects of immunoadsorption thioprine and prednisolone was instituted on were also analysed in relation to clinical fea- day 0. Azathioprine was given at a dose of 3 tures such as age, sex, duration of illness, time mg/kg body weight during the first week and before thymectomy, and thymic histology. 2-5-2 mg/kg body weight for the subsequent The clinical and electrophysiological exam- four weeks (if this dose was not tolerated it inations were performed by the local neuro- was reduced to 2-1P5 mg/kg body weight). logical investigator. The examining Prednisolone was given at a dose of 1-5 mg/kg neurologists were blind as to plasma exchange body weight for the first two weeks, and treatments. All laboratory investigations that 1 5 mg/kg body weight on alternate days would allow the neurological investigator to thereafter. know about the treatment were kept separate Immunoadsorption with IM-TR was car- and were only known to the physician per- ried out on-line with a plasma separator con- forming plasma exchange. Raw data were sisting of a cellulose diacetate membrane. The listed in the study protocol and sent to the blood flow was kept at 70-80 ml/min, and study coordinator for statistical analysis. the transmembrane pressure was less than Because this study was designed as an 40 mm Hg. The flux rate of the filtration (the open, one-armed treatment trial, all compar- flow rate at the IM-TR column) was 20 isons were within groups. Changes from base- ml/min. Adsorption was performed with line observed at 11 days and 28 days after about one plasma volume (an average of 2300 treatment were also compared to analyse the ml of plasma) on days 1, 3, 5, 8, and 10, for a time profile, because the main response vari- total of five treatments. Heparin was used as able in evaluating efficacy was the change in an anticoagulant at an initial dose of 2000 the clinical scores adjusted by the pre-treat- units, then 2000 units/hour during the period ment value. The baseline employed was the of perfusion. Fluids contained in the system observation immediately before the first were reinfused into the patient at the end of immunoadsorption treatment. Statistical the procedure. No plasma proteins were analysis was by Student's t test or x2 test. given. Results are expressed as means (SD). Clinical assessment was based on clinical muscle testing (the myasthenia gravis score of Besinger and Toyka17), the titre of antibodies Results to AChR (a standard immunoprecipitation Pronounced and rapid recovery from myas- assay with solubilised human receptor'), elec- thenic weakness was noted in 11 patients after tromyographic analysis of the neuromuscular a series of immunoadsorption treatments, but transmission block, and side effects if any. in the others the clinical state showed little Briefly, a four-step system for grading muscle change despite a substantial dectease in strength was used (ranging from 0 = normal AChR antibody titre (table 1). Nine of 13 http://jnnp.bmj.com/ to 3 = severe weakness). In each patient, five patients with thymic hyperplasia improved, test items for muscles of the limbs and trunk although only one of six patients with thy- and three test items for the oropharyngeal moma benefited (table 1). The improvement muscles were examined. All items were given was stable in nine of them for at least 32 days the same weight. The myasthenia gravis score after termination of the immunoadsorption. was calculated as the sum of the grades in The clinical effect was significantly correlated each item divided by the number of the items with thymic histology (tables 1 and 2), but no

tested. For each patient, the treatment was correlation was noted with age, sex, duration on September 27, 2021 by guest. Protected copyright. judged to be efficacious if the myasthenia of illness, and time before thymectomy (table gravis score on day 11 was more than 35% 1). Percentage improvement of the mean higher than the pretreatment value. To esti- myasthenia gravis score was 39% on day 1 1 mate the neuromuscular transmission block and 35% on day 28 in patients with thymic quantitatively, the decline in the amplitude of hyperplasia, whereas it was 15% and 1 % in successive responses to repetitive nerve stimu- those with thymoma (table 2). lation at frequencies of 3/s was measured on Serum AChR antibody was reduced by the adductor pollicis muscle and the trapezius about 60% through the treatment of one muscle. It was calculated as a ratio of the fifth plasma volume with IM-TR. No difference response to the first one. These clinical and was noted between the thymic hyperplasia

Table 2 Analysis ofrelation between thymic histology and outcome ofimmunoadsorption therapy % Improvement of Sex Duration Time before Removal rate Mean myasthenia gravis score myasthenia gravis score Thymic No of of illness thymectomy forAChR Ab histology patients Age M F (y) (y) on day 10 (%) Day 0 Day 11 Day 28 Day 11 Day 28 Hyperplasia 13 37(14) 3 10 9(5) 4(3) 64(9) 1-9(03) 1 1(O04)** 1.2(03)** 39(20) 35(19) Thymoma 6 39(15) 2 4 5(5) 3(4) 64(14) 1 9(0 4) 1-7(0-4) 1-7(0 7) 15(20) 11(28) p Value NS NS NS NS NS NS NS <0 05 NS <0 05 <0 05 Values are the means (SD); NS = non-significant; p values are one-tailed. **p<001 vday0. 580 Shibuya, Saw, Osame, Takegami, Doi, Kawanami

Discussion J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.57.5.578 on 1 May 1994. Downloaded from The immunoadsorption together with the 0) 2-5 - 100 immunosuppressive drugs ameliorated myas- o C.) o thenic symptoms in most patients with thymic CO) 2-0 > hyperplasia, and its clinical effects lasted more 0) n than five weeks, but it failed to induce clinical 20 improvement in most patients with thymoma. (a 1 -5 ._ r50 D It is noteworthy that the myasthenia gravis < score, especially for oropharyngeal muscles, CO 1i0 co c remained low even on day 42 despite the < increased AChR antibody titres. 0-5 The patients with thymoma tend to have severe disease, respond poorly to thymec- 0 tomy, show no HLA association, and about 0 4 11 28 42 90% have striated muscle antibodies.23 The Days lack of effect of the immunoadsorption ther- AA A A A apy in cases of thymoma despite a fall in the Immunoadsorption AChR antibodies could be because AChR Mean values ofmyasthenia gravis score and titre ofAChR antibodies in the 11 inlproved antibodies are heterogeneous both with patients after immunoadsorption therapy. Vertical lines are SDs. Consecutive to and immunoadsorption treatments induced a significantfall of the myasthenia gravis score and respect binding specificity biological a decrease in the titre ofAChR antibodies. The myasthenia gravis score remained low even function,2 31819 and because the antibody on day 42 despite the rise in AChR antibody timres. **p < 001 v day 0. response in myasthenia gravis is polyclonal (the antibodies derived from different clones of autoimmune B-lymphocytes have variable and the thymoma cases in the average functional activities).2 There is no identifiable removal rate of AChR antibodies. In E)atients variation in antibody specificity that accounts with remission after the treatment, the AChR for variation in clinical severity.'819 It seems antibodies were reduced by 58 (12)'% (n = likely that there are endogenous factors affect- 53; range 33-75%) similar to the average of ing the safety margin for neuromuscular 61 (13)% (n = 38; range 35-85%) in E)atients transmission and there are different immuno- who showed no change. Albumin fell byy only 6 logical and genetic factors that control differ- (3)% (n = 50). ences in a patients' phenotype. Resistance to In the 1 patients who improved, r(ecovery plasma exchange may also be due to irre- usually occurred within 48 hours after tthe first versible changes in muscle as a consequence adsoption, reaching its peak one to foiur days of the severity of the disease or to degenera- after the last adsorption. The total mya[sthenia tion of the postsynaptic membrane due to gravis score (five items for muscles oef limbs long-term anticholinesterase therapy.20 and trunk, and three items for oropharyngeal Dau8 reported that the factors correlating muscles) was significantly reduced fr4 om 19 with the best clinical response were short (0.3) (n = 11) on day 0 to 0-9 (0-2) on day duration of illness, male sex of the patient, 11; to 1 0 (0 3) on day 28; and to 1-2 (I0 4) on and treatment with both prednisolone and day 42 (p < 0 01). The mean myatsthenia azathioprine during plasma exchange. We http://jnnp.bmj.com/ gravis score for oropharyngeal rnuscles could not, however, find a clear relation showed the lowest value on day 28 ancI it was between the effectiveness of the immuno- still low on day 42. The titre of AChR anti- adsorption combined with prednisolone and body was 44% on day 14 and 65% on (day 28, azathioprine and the duration of the illness, in reference to the value on day 0 taiken as an age and sex difference, or the timing of the 100% (figure). These values were signi:ficantly thymectomy.

different from the titre of antibody on dlay 0 (p With immunoadsorption therapy using on September 27, 2021 by guest. Protected copyright. < 0 01). The AChR antibodies gradually IM-TR, the AChR antibodies can be reduced increased after the adsorption th( erapies, as effectively as with conventional plasma whereas the myasthenia gravis scores exchange using centrifugation," without loss remained low (figure). of albumin and no serious complications."1'6 Analysis of the neuromuscular transrmission The titre of the AChR antibodies began to rise block was performed in nine of 11 rtemitted soon after each immunoadsorption treatment, patients after the treatment. In the evoked but its rise was slow and its titre was less than electromyogram the decline of the fifth 80% of the pretreatment values even 30 days response compared with the first was signifi- after the treatments. An inverse association of cantly improved in the adductor polliciis mus- the clinical state with AChR antibody titres is cle from 33 (13)% on day 0 to 13 (5)% on day usually seen just after a series of plasma 28 (p < 0.01). The mean recovery raite was exchanges.67 Rapid removal of the circulating 59%. The degree of decline in amplitude on antibodies is followed by their redistribution, the trapezius muscle was also signilficantly however,6 and may affect the rate of synthesis reduced from 51% to 35% (p < 0-01). In lOO by removing inhibitory feedback2' or causing a procedures, the treatment caused no serious reduction of catabolism combined with an complications. There were eight mild unchanged rate of synthesis.22 A rebound hypotensive reactions and five patienits had increase in the AChR antibody usually occurs mild symptoms such as nausea, vomitiu ng, and within seven days, and the original titre is headache, which can accompany any type of restored within 14 days67 with accompanying extracorporeal circulation. clinical worsening unless azathioprine, which Immunoadsorption therapyfor myasthenia gravis 581

have a action on clones of the 8 Dau PC. Response to plasmapheresis and immunosup- J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.57.5.578 on 1 May 1994. Downloaded from may cytotoxic pressive drug therapy in sixty myasthenia gravis patients. antigen,6 and prednisolone, *hich may sup- Ann N YAcad Sci 1981;377:700-8. antibody from B-cells,26 are 9 Consensus Development Conference, National Institute of press production Health. The utility of therapeutic plasmapheresis for given. Azathioprine and prednisolone are use- neurological disorders. JAMA 1986;256: 1333-7. in the treatment of but 10 Bussel A, Sitthy X, Reviron J. Technical aspects and com- ful myasthenia gravis, plications ofplasma-exchange. La Ricerca Clin Lab 1983; slow in action, and if given alone the average 13:111-32. required for a 50% reduction in titre is 11 Yamazaki Z, Fujimori Y, Takahara T, et al. Efficiency and time biocompatibility of a new immunosorbent. Trans Am Soc three to five months,2324 whereas plasma Artificial Intern Organs 1982;28:318-23. alone can reduce more than 12 Sato T, Nishimiya J, Arai K, et al. Selective removal of exchange rapidly anti-acetylcholine receptor antibodies in sera from 60% of the circulating antibodies. Our results patients with myasthenia gravis in vitro with a new showed that a combination of immunoad- immunosorbent. In: Oda T, ed. Therapeutic plasmaphere- sis (III). Stuttgart: Schattauer, 1983:565-8. sorption therapy and immunosuppressive 13 Shibuya N, Nagasato K, Kinoshita N, et al. medication is to maintain the bene- Immunoadsorbent perfusion therapy in patients with necessary myasthenia gravis. In: Shiokawa Y, Inoue N, eds. ficial effects and to avoid a rebound increase Current practice in therapeutic plasmapheresis. Amsterdam: in the AChR antibodies caused by phenom- Excerpta Medica 694;1985:166-72. 14 Shibuya N, Nagasato K, Shibayama K, et al. ena induced by IgG depletion.2122 Immunoadsorption therapy in neurologic diseases; It was concluded that combined myasthenia gravis, multiple sclerosis and Guillain-Barre syndrome. In: Oda T, et al, eds. Proceedings of the first immunoadsorption therapy and immunosup- international congress ofthe world apheresis association: ther- pressive drug treatment is useful in controlling apeutic plasmapheresis (VI). Cleveland: ISAO Press 1987;311:122-8. symptoms in patients with severe myasthenia 15 Heininger K, Hendricks M, Toyka KV. Myasthenia gravis: gravis who are otherwise unresponsive. a new semiselective procedure to remove acetylcholine receptor autoantibodies from plasma. Plasma Ther Transfus Technol 1985;6:771-5. 16 Avanzi G, Marconi G, Calacoci L, et al. Plasmafiltration This study was carried out as a joint Japanese-German med- and immunoadsorption on a tryptophane column in ical project on immunoadsorption therapy for myasthenia myasthenia gravis: a study of 22 patients. Ital J Neurol gravis. A part of this study was presented at the 3rd Sci 1991;1(suppl):47-51. International Congress of the World Apheresis Association in 17 Besinger UA, Toyka KV, Homberg M, et al. Myasthenia Amsterdam, The Netherlands, April 1990. The work was sup- gravis: long-term correlation of binding and bungaro- ported by a grant from the Japan Health Sciences Foundation. toxin blocking antibody against acetylcholine receptors with changes in disease severity. Neurology 1983;33: 1316-21. 18 Tzartos SJ, Seybold ME, Lindstrome JM. Specificities of antibodies to acetylcholine receptors in sera from myas- 1 Lindstrom JM, Seybold ME, Lennon VA, et al. Antibody thenia gravis patients measured by monoclonal antibod- to acetylcholine receptor in myasthenia gravis: preva- ies. Proc Nad Acad Sci USA 1982;79:188-92. lence, clinical correlates, and diagnostic value. Neurology 19 Mittag T, Massa T, Komfeld P, et al. Multiple forms of 1976;26: 1054-9. antiacetylcholine receptor antibody in myasthenia gravis. 2 Engel AG. Acquired autoimmune myasthenia gravis. In: Muscle Nerve 1981;4:16-25. Engel AG, Banker BQ, eds. Myology. New York: 20 Engel AG, Lambert EH, Santa T. Study of long-term anti- McGraw-Hill, 1986:1925-54. cholinesterase therapy: effects on neuromuscular trans- 3 Vincent A, Newsom-Davis J. Anti-acetylcholine receptor mission and on motor end-plate fine structure. Neurology antibodies. J Neurol Neurosurg Psychiatry 1980;43: 1973;23: 1273-81. 590-600. 21 Sturgill BC, Worniak MJ. Stimulation of proliferation of 4 Drachman DB, Angus CW, Adams RN, et al. Myasthenic 19S antibody-forming cells in the spleens of immunized antibodies cross-link acetylcholine receptors to acceler- guinea pigs after exchange transfusion. Nature 1970;228: ate degradation. N EnglJ Med 1978;298: 1116-22. 1304-5. 5 Engel AG. Morphologic and immunopathologic findings 22 Charlton B, Schindhelm K. The effect of extracorporeal in myasthenia gravis and in congenital myasthenic syn- antibody removal on antibody synthesis and catabolism dromes. J Neurol Neurosurg Psychiatry 1980;43:577-89. in immunized rabbits. Clin Exp Immunol 1985;60: 6 Dau PC, Lindstrom JM, Cassel JK, et al. Plasmapheresis 457-64. and immunosuppressive drug therapy in myasthenia 23 Tindal RSA. Humoral immunity in myasthenia gravis: gravis. NEnglJMed 1977;297:1134-40. effects of steroids and thymectomy. Neurology 1980;30: http://jnnp.bmj.com/ 7 Newsom-Davis J, Pinching AJ, Vincent A, et al. Function 554-7. of circulating antibody to acetylcholine receptor in myas- 24 Mertens HG, Hertel G, Reuther P, et al. Effects of thenia gravis: investigation by plasma exchange. immunosuppressive drugs (Azathioprine). Ann N Y Neurology 1987;28:266-72. Acad Sci 1981;377:691-9. on September 27, 2021 by guest. Protected copyright.