- Email: [email protected]
Clinical heterogeneity of seronegative myasthenia gravis
Neuromusc.Disord.,Vol. 6. No. 3, pp. 155-161, 1996 Copyright ~ 1996 ElsevierScience Ltd. All rights reserved Printed in Great Britain 0960-8966/96 515.00 + .00
Pergamon
PII: S0960-8966(96)00009-0 CLINICAL
HETEROGENEITY
OF SERONEGATIVE
MYASTHENIA
GRAVIS A. EVOLI*+, A. P. BATOCCHI*, M. LO MONACO*, S. SERVIDEI*, L. PADUA*, L. MAJOLINI* and P. TONALI*~ *Institute of Neurology, Catholic University, Rome, Italy; +*CSS Hospital IRCCS, San Giovanni Rotondo, Italy
(Received 4 October 1995; revised 27 November 1995; accepted 18 January 1996)
Abstract--We studied 38 patients affected by seronegative myasthenia gravis (SNMG) with age at the onset of the disease ranging from 6 to 66 yr. Clinical follow-up lasted at least 2 yr. Patients' lymphocyte cultures showed in no case anti-acetylcholine receptor antibody production; HLA associations did not differ significantly from those in seropositive MG. In most cases (33 out of 38) the disease differed from seropositive MG only in the low incidence of thymic pathology. In five cases the clinical picture was characterized by prevalent involvement of ocular and bulbar muscles and by a more stable course. These last group of patients did not respond satisfactorily to immunosuppressants, but they clearly improved with plasma-exchange. Mice injected with plasma IgG from four patients (two with typical clinical picture and two with prevalent oculobulbar involvement) showed a defect of neuromuscular transmission. In our experience, SNMG is a heterogeneous disease. A humorally mediated pathogenesis appears to be operating in both groups of patients we described. Copyright © 1996 Elsevier Science Ltd. Key words: Myasthenia gravis, seronegative myasthenia gravis, anti-AChR antibody,
thymectomy. INTRODUCTION
The hallmark of myasthenia gravis (MG) is the loss of acetylcholine receptors (AChRs) at the motor end-plate caused by specific IgG autoantibodies (abs) [1]. These immunoglobulins are detectable by radioimmunoassay in about 60% of patients with ocular myasthenia and in 90% of cases with generalized disease [2]. The term 'seronegative M G ' (SNMG) is commonly applied to generalized myasthenia with persistently negative results of anti-AChR ab titration [3, 4]. There is evidence that autoimmune mechanisms are operating also in this form: immunosuppressive therapy and plasma-exchange improve the disease course [5-7] and patients' plasma or IgG can transfer the neuromuscular transmission defect to mice [4, 5, 8]. However, it is still under discussion as to whether S N M G is to be considered merely one end of the M G spectrum [9] or whether, alternatively, it is a distinct entity in which
pathogenetic antibodies are directed towards antigens different from AChR [5]. Moreover, it has recently been suggested that SNMG could be a heterogeneous disease in which different target antigens might be involved [4, 8] and plasma factors other than IgG could be operating [3, 4]. With respect to clinical features, Birmanns et aL [10] identified two subgroups of S N M G patients and suggested that some cases unresponsive to immunosuppressive drugs could be affected by a non-immune-mediated disease. As a consequence of these unresolved questions, it is not clear at present if all therapeutic strategies used in seropositive M G (SPMG) are indicated also for these patients. We herewith report clinical and laboratory data of a series of 38 S N M G patients and provide further evidence that circulating factors are responsible for the disease. PATIENTS AND M E T H O D S
-tAuthor to whom correspondence should be addressed at: Institute of Neurology, Catholic University, Largo A. Gemelli, 8, 00168 Rome, Italy.
Out of 460 patients with generalized MG, 48 (10.5%) resulted seronegative. Anti-AChR ab 155
156
A. Evoli et al.
assay was performed according to Lindstrom [l 1], with minor modifications [12] using human muscle AChR as antigen. For each patient, seronegativity was confirmed by at least three ab titrations with different AChR extracts. Antibodies against the toxin-binding sites on AChR were tested according to Besinger et al. [13] and were negative in all these cases. Thirty-eight patients were included in this study: in all of them seronegativity was assessed before immunosuppressive therapy and the clinical follow-up lasted at least 2 yr. Ten patients were excluded for the following reasons: corticosteroid treatment at the time of the first ab titration (seven cases), insufficient follow-up data (three cases). The diagnosis of MG was based on the association of typical history and signs together with at least one of these features: positive response to edrophonium i.v. and abnormal decrement (>11%) of the third to fifth compound muscle action potential (CMAP) on low-rate repetitive nerve stimulation (RNS) or increased jitter on single fibre-EMG (SFEMG). Electrophysiologic investigations, in these patients, also included conventional EMG, high-rate (50 Hz) RNS and nerve conduction studies. MG severity was graded according to Osserman [14]. Pyridostigmine was used in all cases; prednisone was administered according to a previously described schedule [15] to 28 patients with severe MG symptoms, poorly controlled by anticholinesterase (ACHE) drugs; azathioprine was associated in eight cases; 10 patients received one or more courses of plasma-exchange; thymectomy was performed in 29 patients, in most cases before immunosuppressive therapy. The response to treatment was considered good when the patients' clinical status, at the end of follow-up, was classified as remission (patient asymptomatic without any therapy for at least 6 months), pharmacological remission (patient asymptomatic but still taking therapy) or marked improvement (stable, marked improvement of myasthenic signs with reduction in drug dosage). HLA typing was carried out in 18 SNMG patients using 11th Histocompatibility Workshop-validated antisera in complementmediated micro-cytotoxicity tests. Antigen
frequencies were compared with those in 49 patients affected with generalized SPMG and in 392 healthy Italian subjects (from Red Cross Immunohematology Department, Rome). In vitro production of anti-AChR abs was tested in the same 18 SNMG patients, in 10 SPMG cases and in six normal controls. Most patients were under long-term corticosteroid/immunosuppressive therapy when bled. PBL were isolated by isopaque-Ficoll density gradient and resuspended in RPMI-1640 medium plus 10% fetal calf serum and antibiotics. Triplicated cultures of 1× 106 lymphocytes were performed, with and without 10 gg ml 1 PWM, for 7 days at 37°C in humidified air with 5% CO2. PWM-driven anti-AChR ab production was assayed in cell-free supernatants according to the method described by Kuks et al. [161. Passive transfer studies were performed using plasma Ig from four seronegative, three seropositive MG patients (positive control) and healthy donors. Ammonium sulphate-precipitated IgG [5] were injected intraperitoneally (i.p.) into female C57 BL6 mice (18-20 g) at a dose of 2 mg g b.w. -l day-1 for 4 consecutive days. On the fifth day from the start of injection, electrophysiologic studies were performed in vivo as follows: under penthobarbital anaesthesia (40-60 lag g b.w. i i.p.-l), the sciatic nerve was stimulated at the notch using two monopolar needle electrodes and the CMAP of the tibialis anterior muscle was recorded by a monopolar needle electrode inserted subcutaneously at the muscle belly and referred to the paw. The electrical stimulus was a square pulse of 0.05 ms duration with an intensity twice that required to produce the maximum response, of CMAP. The nerve was stimulated repetitively at 3 Hz before and after injection of d-tubocurarine (0.15 gg g b.w. -I i.p. 1). Amplitute variation of CMAP negative peak occurring during the first three to six responses of each train of stimuli was calculated and in case of significant decremental pattern (CMAP amplitute reduction >10%) correction with edrophonium (2 gg i.p.) was attempted. A group of three mice was injected with each Ig preparation. Limb muscle (deltoid or biceps brachialis) biopsies were obtained from four SNMG patients and routine morphological studies were performed. Statistical analysis was performed by Z2 test.
Seronegative Myasthenia Gravis RESULTS
The series of SNMG patients was made up of 12 males and 26 females (M:F = 1:2.16) with age at disease onset ranging from 6 to 66 yr (mean 32.3). All patients but one had a positive edrophonium test; electrophysiological signs of neuromuscular exaustion (decremental pattern or increased jitter) were present in all cases; electromyographic features of the Lambert-Eaton myasthenic syndrome and repetitive response to single nerve stimuli were absent. In no case was there parents' consanguinity. HLA associations
We did not find any significant difference between patients with seropositive and seronegative generalized MG. The frequencies of B8, DR3 and DQ2 antigens in SNMG were similar to those of controls, while they were significantly higher in SPMG; HLA-DQ3 showed an increased frequency (P<0.01) both in seropositive and in seronegative M G when compared to the general population; HLA-DR2 was weakly associated with SNMG. These data are shown in Table 1. The haplotype B8 DR3 DQ2 was present only in one out of 18 SNMG cases. Clinical data
In 33 of our 38 patients the clinical picture was indistinguishable from SPMG: weakness distribution was typical and daily fluctuations of symptoms were evident. The patients were nine males and 24 females with age at onset ranging from 13 to 66 yr (mean 36.2). Eighteen patients (54%) were affected by mild disease (group 2A according to Osserman), 15 had severe symptoms (groups 2B, 3 and 4). Twentythree patients had undergone thymectomy; the pathological examination of the thymus
157
revealed a normal/involuted thymus in 17 patients (74%), thymic hyperplasia in five (21.7%) and a small nodular thymoma in one (4.3%). An associated autoimmune disease was present in three cases. All patients showed a good response to pyridostigmine, immunosuppressive therapy and plasma-exchange. The efficacy of thymectomy is difficult to assess as it was associated with immunosuppressive therapy in most cases. Plasma IgG from two of these patients transferred the disease to mice (see below). In five patients the clinical picture was characterized by the prevalent involvement of ocular and bulbar districts. All these patients showed opthalmoparesis, marked facial weakness, nasal speech, mild dysphagia and suffered from one or more respiratory crises. Limb muscles were less severely affected, but they also became clearly weakened during exacerbation phases. Clinical course was more stable than in typical MG and, in particular, daily fluctuations were scarcely evident. This group consisted of three males and two females with age at onset ranging from 6 to 23 years (mean 14.8). All patients underwent thymectomy; the histological examination of the thymus showed in no case germinal centres in the medulla. Other autoimmune diseases were not present, but the father of one of these patients was affected by SPMG [17]. The clinical response to pyridostigrnine, although clearly positive in the early stages of the disease, was generally poor with muscarinic side-effects even with small doses; so pyridostigmine was finally withdrawn in three cases. All patients received prednisone with mild benefit and are still under treatment; azathioprine had been administered for at least 1 yr to three patients and then withdrawn as ineffective; thymectomy did not appear to influence the course of the disease. Four patients were submitted to plasma-exchange with deft-
Table 1. HLA associations in seronegative generalized MG, seropositive generalized MG and controls Seronegative MG (n = 18) HLA antigen AI B8 DR2 DR3 DQ2 DQ3
No. 5 2 6 5 6 9*
*Significantly different from controls.
% 27.7 11 33 27.7 33 50
Seropositive MG (n = 49) No. 18 17" 11 23* 26* 18"
Controls (n = 392)
%
No.
%
36.7 34.7 22.4 46.9 53 36.7
102 46 66 77 133 72
26 11.7 16.8 19.6 33.9 18.4
158
A. Evoli et aL
nite improvement, two of them received several plasma-exchange courses which were followed by transient improvement of symptoms. Plasma IgG from two of these patients, when injected into mice, transferred the defect of neuromuscular transmission (see below). Muscle biopsies, from four patients with prevalent oculobulbar symptoms, showed increased fiber diameter variability, scattered focal necrosis and single fiber atrophy mainly affecting type 2; moreover, slight type 1 predominance was present in two of these cases (Fig. 1). In vitro anti-AChR ab production Lymphocyte cultures were performed with PBL from 18 SNMG patients, five of whom were affected by prevalent oculobulbar symptoms. In vitro anti-AChR ab production was never detected in cultures from SNMG patients and controls, while anti-AChR abs were measurable in culture supernatants from five out of 10 SPMG cases. These findings are shown in Fig. 2. Passive transfer studies
Plasma IgG of SNMG patients were from two patients with typical clinical picture and from two with prevalent oculobulbar involvement. RNS induced a significant decremental pattern in mice injected with IgG from SPMG and SNMG patients. The decremental response
occurred after d-tubocurarine injection and was corrected by edrophonium. Figure 3 shows the electrophysiological records in a mouse injected with plasma IgG from a patient with oculobulbar symptoms. No difference was found between mice treated with IgG from SPMG and SNMG patients with respect to the number of mice affected in each group and the severity of decremental response. Neither weakness nor electrophysiological signs of neuromuscular transmission impairment were observed in mice injected with IgG from healthy donors. DISCUSSION
Pathogenetic mechanisms of SNMG are still under debate. Negative results of the anti-AChR assay could be due merely to technical reasons such as too low an ab level or loss of important antigenic determinants in receptor solubilization [6, 9]. Alternatively, the target antigen could be different from AChR. This last hypothesis is supported by those studies reporting lack of correlation between impairment of neuromuscular transmission and end-plate AChR reduction both in intercostal muscle biopsies from SNMG patients and in mice submitted to passive transfer [4, 5]. More recently, Burges et al. have demonstrated that a pre-synaptic defect (reduction in quantal content) is operating in some of these cases [8]. Moreover, the effector mechanisms might be mediated by serum components other than IgG
Fig. 1. (A) HE; (B) ATP-ase pH 4.3. Deltoid biopsy from a SNMG patient with prevalent oculobulbar symptoms. Increased fibre diameter variability and single fibre atrophy (A). Type 1 fibre predominance (B).
Seronegative Myasthenia Gravis anti-AChR ab (fmolll0'
cells)
8-
0
•
•
o o o o
•
•
•
o
o
o
•
•
•
•
•
•
I
nnUili
I
Seronegative MG Seropositive MG
•
•
•
i
t
Controls
Fig. 2. P W M - d r i v e n in vitro synthesis of a n t i - A C h E antibodies.
such as high molecular weight molecules possibly IgM [18], and non-antibody serum factors co-purifying with the IgG fraction [3]. A spontaneous synthesis of anti-AChR ab by PBL cultures implies the presence of auto-reactive B cells. Such a production, under PWM stimulation, has been demonstrated in 50-76% of MG patients [19-21]. In our experience antiAChR abs were undetectable in culture supernatants from SNMG patients. These data are in agreement with previous reports [10, 20] and confirm the results of serum assays. To investigate pathogenetic aspects of SNMG we have compared HLA frequencies in these patients with those in seropositive cases and in the general population. It is well known that, in Caucasians, early-onset non-thymoma MG is strongly associated with HLA-B8, DR3 and DQ2 [19]. In seronegative patients the frequencies of these antigens were not increased, despite the young age at onset. We did not find any significant difference in HLA associations between SPMG and SNMG. HLA-DQ3 occurred at increased frequency in both MG series when compared with the general Italian population. The significance of this association is unknown. A common feature of SNMG is the low frequency of thymic pathology and, in particular, the absence of thymoma [3, 7, 10, 19]. In our experience the association with
159
thymoma, although very rare, can occur; the frequencies of both thymic hyperplasia and involution in our patients are in accordance with other reports. The role of the thymus in SNMG is at present far from clear. Willcox et al. [20] described sparse germinal centers and expanded medullary T-cell areas as distinctive thymus abnormalities in seronegative patients. The authors suggested these changes could be secondary also because thymic cultures from the same patients failed to produce antiAChR antibodies. Thymectomy has generally been reported as scarcely effective in these cases [3, 7, 10, 19, 20] and, consequently, its role in the treatment of SNMG is highly controversial. Clinical heterogeneity of SNMG has been reported by Birmanns et al. [10]: these authors described two forms of the disease and suggested that oculobulbar SNMG which was unresponsive to immunosuppressants could be a non-immune-mediated disorder. Clinical characteristics in our series were similar to those described but all our patients were affected with generalized disease and oculobulbar cases had severe symptoms with respiratory crises. In most of our cases SNMG appears clinically indistinguishable from the seropositive disease; moreover, the efficacy of plasmaexchange and immunosuppressive drugs as well as the results of passive transfer studies leave little doubt that it is an autoimmune disorder. In as much as thymic abnormalities are uncommon in SNMG, the only thymoma and the few thymic hyperplasias were found among these cases. These features suggest that in some patients the disease could be the end of the MG spectrum and that anti-AChE abs could be present at undetectable levels; on the other hand, antigens different from AChR could be involved in other cases. Further clinical experience is necessary to define the therapeutic role of thymectomy. In patients with prevalent oculobulbar involvement the course of the disease is more stable than typical MG and the response to immunosuppressive therapy is less satisfactory. Moreover, thymic abnormalities were absent in every case. In these patients a congenital myasthenia should first be ruled out. The onset of symptoms after the first years of life and the negative family history favor an acquired disease. Muscle biopsy studies showed abnor-
160
A. Evoli et aL
RNS at 3 Hz
To*
TIo
TI7 T3s
TJe
Fig. 3. Decremental response in a mouse injected with plasma IgG from a SNMG patient with prevalent oculobulbar symptoms. *d-Tubocurarine injection; **edrophonium injection. T indicates minutes after injection; calibration bar = 10 inV. Twenty-nine minutes after the injection of d-tubocararine, RNS elicited a significant decremental response which was corrected by the injection of edrophonium.
malities such as focal necrosis and single fiber atrophy which have been described also in muscles from SPMG patients [22]. In two cases these changes were associated with mild type-1 fiber predominance which is a common although not a specific finding in congenital myasthenic syndromes [23]. Immunocytochemical and in vitro electrophysiological studies are necessary to fully investigate the underlying defect of neuromuscular transmission. Nevertheless, the efficacy of plasma-exchange and passive transfer experiments strongly suggest that these cases also are affected by a
humorally mediated disorder. Acknowledgements The financial support from Telethon (grant No. 400) is gratefully acknowledged.
REFERENCES
1. Engel AG. Myasthenia gravis and myasthenic syndromes. Ann Neurol 1984; 16: 519-534. 2. Vincent A, Newsom-Davis J. Acetylcholine receptor antibody as a diagnostic test for myasthenia gravis: results in 153 validated cases and 2967 diagnostic assays. J Neurol Neurosurg Psychiatry 1985; 48: 1246-1252. 3. Verma P K, Oger J J F. Seronegative generalized myasthenia gravis: low frequency of thymic pathology. Neurology 1992; 42: 586-589. 4. Vincent A, Li Z, Hart A, et aL Seronegative myasthenia gravis. Evidence for plasma factor(s) interfering with acetylcholine receptor function. Ann N Y Acad Sci 1993; 681: 529-538. 5. Mossman S, Vincent A, Newsom-Davis J. Myasthenia gravis without acetylcholine receptor antibody: a distinct disease entity. Lancet 1986; i: 116-119. 6. Soliven B C, Lange D J, Penn A S, et al. Seronegative myasthenia gravis. Neurology 1988; 38:514-517. 7. Evoli A, Bartoc~ioni E, Batoc~hi A P, Scuderi F, Tonali P. Anti AChR-negative myasthenia gravis:
Seronegative Myasthenia Gravis
8.
9. 10.
11.
12. 13.
14. 15.
clinical and immunological features. Clin Invest Med 1989; 12: 104-109. BurgesJ, Vincent A, Molenaar P C, Newsom-Davis J, Peer C, Wray D. Passive transfer of seronegative myasthenia gravis to mice. Muscle Nerve 1994; 17: 1393-1400 Drachman D B, De Silva S, Ramsay D, Pestronk A. Humoral pathogenesis of myasthenia gravis. Ann N Y Acad Sci 1987; 505: 90-105. Birmanns B, Brenner T, Abramsky O, Steiner I. Seronegative myasthenia gravis: clinical features, response to therapy and synthesis of acetylcholine receptor antibodies in vitro. J Neurol Sci 1991; 102: 184-189. Lindstrom J. An assay for antibody to human acetylcholine receptor in serum from patients with myasthenia gravis. Clin Immunol Immunopathol 1977; 7: 36-43. Bartoccioni E, Scuderi F, Scoppetta C, et al. Myasthenia gravis, thymectomy and anti-acetylcholine receptor antibody. J Neurol 1980; 224: 9-15. BesingerU A, Toyka K V, Homberg M, Heininger K, Hohlfeld R, Fateh-Moghadam A. Myasthenia gravis: long-term correlation of binding and bungarotoxin blocking antibodies against acetylcholine receptor with changes in disease severity. Neurology 1983; 33:1316-1321. Osserman K E, Genkins G. Studies on myasthenia gravis: review of a twenty-year experience over 1200 patients. Mt Sinai J Med 1971; 38: 497-537. Evoli A, Batocchi A P, Palmisani M T, Lo Monaco M, Tonali P. Long-term results of corticosteroid therapy in myasthenia gravis. Eur Neurol 1992; 32: 37-43.
161
16. Kuks J B M, Limburg P C, Oosterhuis H J G H, The T H. Antibodies to acetylcholine receptor in myasthenia gravis. In vitro synthesis by peripheral blood lymphocytes before and after thymectomy. Clin Exp Immunol 1992; 87: 246-250. 17. Provenzano C, Arancio O, Evoli A, et al. Familial autoimmune myasthenia gravis with different pathogenetic antibodies. J Neurol Neurosurg Psychiat 1988; 51: 1228-1230. 18. Yamamoto T, Vincent A, Ciulla T A, Lang B, Johnston I, Newsom-Davis J. Seronegative myasthenia gravis: a plasma factor inhibiting agonist-induced acetylcholine receptor function copurifies with IgM. Ann Neurol 1991; 30: 550-557. 19. Newsom-Davis J, Willcox N, Schluep M, et al. Immunological heterogeneity and cellular mechanisms in myasthenia gravis. Ann N Y Acad Sci 1987; 505: 12-26. 20. WillcoxN, Schluep M, Ritter M A, Newsom-Davis J. The thymus in seronegative myasthenia gravis. J Neurol 1991; 238: 256-261. 21. Lisak R P, Levinson A I, Zweiman B, Kornstein M J. In vitro synthesis of IgG and antibodies to AChR by peripheral and thymic lymphocytes. Ann N Y Acad Sci 1987; 505: 39-49. 22. Oosterhuis H J G H. Myasthenia gravis. In Clinical Neurology and Neurosurgery Monographs (Vol. 5). Churchill Livingstone, 1984. 23. Engel A G. The investigation of congenital myasthenic syndromes. Ann N Y Acad Sci 1993; 681: 425-434.
Pergamon
PII: S0960-8966(96)00009-0 CLINICAL
HETEROGENEITY
OF SERONEGATIVE
MYASTHENIA
GRAVIS A. EVOLI*+, A. P. BATOCCHI*, M. LO MONACO*, S. SERVIDEI*, L. PADUA*, L. MAJOLINI* and P. TONALI*~ *Institute of Neurology, Catholic University, Rome, Italy; +*CSS Hospital IRCCS, San Giovanni Rotondo, Italy
(Received 4 October 1995; revised 27 November 1995; accepted 18 January 1996)
Abstract--We studied 38 patients affected by seronegative myasthenia gravis (SNMG) with age at the onset of the disease ranging from 6 to 66 yr. Clinical follow-up lasted at least 2 yr. Patients' lymphocyte cultures showed in no case anti-acetylcholine receptor antibody production; HLA associations did not differ significantly from those in seropositive MG. In most cases (33 out of 38) the disease differed from seropositive MG only in the low incidence of thymic pathology. In five cases the clinical picture was characterized by prevalent involvement of ocular and bulbar muscles and by a more stable course. These last group of patients did not respond satisfactorily to immunosuppressants, but they clearly improved with plasma-exchange. Mice injected with plasma IgG from four patients (two with typical clinical picture and two with prevalent oculobulbar involvement) showed a defect of neuromuscular transmission. In our experience, SNMG is a heterogeneous disease. A humorally mediated pathogenesis appears to be operating in both groups of patients we described. Copyright © 1996 Elsevier Science Ltd. Key words: Myasthenia gravis, seronegative myasthenia gravis, anti-AChR antibody,
thymectomy. INTRODUCTION
The hallmark of myasthenia gravis (MG) is the loss of acetylcholine receptors (AChRs) at the motor end-plate caused by specific IgG autoantibodies (abs) [1]. These immunoglobulins are detectable by radioimmunoassay in about 60% of patients with ocular myasthenia and in 90% of cases with generalized disease [2]. The term 'seronegative M G ' (SNMG) is commonly applied to generalized myasthenia with persistently negative results of anti-AChR ab titration [3, 4]. There is evidence that autoimmune mechanisms are operating also in this form: immunosuppressive therapy and plasma-exchange improve the disease course [5-7] and patients' plasma or IgG can transfer the neuromuscular transmission defect to mice [4, 5, 8]. However, it is still under discussion as to whether S N M G is to be considered merely one end of the M G spectrum [9] or whether, alternatively, it is a distinct entity in which
pathogenetic antibodies are directed towards antigens different from AChR [5]. Moreover, it has recently been suggested that SNMG could be a heterogeneous disease in which different target antigens might be involved [4, 8] and plasma factors other than IgG could be operating [3, 4]. With respect to clinical features, Birmanns et aL [10] identified two subgroups of S N M G patients and suggested that some cases unresponsive to immunosuppressive drugs could be affected by a non-immune-mediated disease. As a consequence of these unresolved questions, it is not clear at present if all therapeutic strategies used in seropositive M G (SPMG) are indicated also for these patients. We herewith report clinical and laboratory data of a series of 38 S N M G patients and provide further evidence that circulating factors are responsible for the disease. PATIENTS AND M E T H O D S
-tAuthor to whom correspondence should be addressed at: Institute of Neurology, Catholic University, Largo A. Gemelli, 8, 00168 Rome, Italy.
Out of 460 patients with generalized MG, 48 (10.5%) resulted seronegative. Anti-AChR ab 155
156
A. Evoli et al.
assay was performed according to Lindstrom [l 1], with minor modifications [12] using human muscle AChR as antigen. For each patient, seronegativity was confirmed by at least three ab titrations with different AChR extracts. Antibodies against the toxin-binding sites on AChR were tested according to Besinger et al. [13] and were negative in all these cases. Thirty-eight patients were included in this study: in all of them seronegativity was assessed before immunosuppressive therapy and the clinical follow-up lasted at least 2 yr. Ten patients were excluded for the following reasons: corticosteroid treatment at the time of the first ab titration (seven cases), insufficient follow-up data (three cases). The diagnosis of MG was based on the association of typical history and signs together with at least one of these features: positive response to edrophonium i.v. and abnormal decrement (>11%) of the third to fifth compound muscle action potential (CMAP) on low-rate repetitive nerve stimulation (RNS) or increased jitter on single fibre-EMG (SFEMG). Electrophysiologic investigations, in these patients, also included conventional EMG, high-rate (50 Hz) RNS and nerve conduction studies. MG severity was graded according to Osserman [14]. Pyridostigmine was used in all cases; prednisone was administered according to a previously described schedule [15] to 28 patients with severe MG symptoms, poorly controlled by anticholinesterase (ACHE) drugs; azathioprine was associated in eight cases; 10 patients received one or more courses of plasma-exchange; thymectomy was performed in 29 patients, in most cases before immunosuppressive therapy. The response to treatment was considered good when the patients' clinical status, at the end of follow-up, was classified as remission (patient asymptomatic without any therapy for at least 6 months), pharmacological remission (patient asymptomatic but still taking therapy) or marked improvement (stable, marked improvement of myasthenic signs with reduction in drug dosage). HLA typing was carried out in 18 SNMG patients using 11th Histocompatibility Workshop-validated antisera in complementmediated micro-cytotoxicity tests. Antigen
frequencies were compared with those in 49 patients affected with generalized SPMG and in 392 healthy Italian subjects (from Red Cross Immunohematology Department, Rome). In vitro production of anti-AChR abs was tested in the same 18 SNMG patients, in 10 SPMG cases and in six normal controls. Most patients were under long-term corticosteroid/immunosuppressive therapy when bled. PBL were isolated by isopaque-Ficoll density gradient and resuspended in RPMI-1640 medium plus 10% fetal calf serum and antibiotics. Triplicated cultures of 1× 106 lymphocytes were performed, with and without 10 gg ml 1 PWM, for 7 days at 37°C in humidified air with 5% CO2. PWM-driven anti-AChR ab production was assayed in cell-free supernatants according to the method described by Kuks et al. [161. Passive transfer studies were performed using plasma Ig from four seronegative, three seropositive MG patients (positive control) and healthy donors. Ammonium sulphate-precipitated IgG [5] were injected intraperitoneally (i.p.) into female C57 BL6 mice (18-20 g) at a dose of 2 mg g b.w. -l day-1 for 4 consecutive days. On the fifth day from the start of injection, electrophysiologic studies were performed in vivo as follows: under penthobarbital anaesthesia (40-60 lag g b.w. i i.p.-l), the sciatic nerve was stimulated at the notch using two monopolar needle electrodes and the CMAP of the tibialis anterior muscle was recorded by a monopolar needle electrode inserted subcutaneously at the muscle belly and referred to the paw. The electrical stimulus was a square pulse of 0.05 ms duration with an intensity twice that required to produce the maximum response, of CMAP. The nerve was stimulated repetitively at 3 Hz before and after injection of d-tubocurarine (0.15 gg g b.w. -I i.p. 1). Amplitute variation of CMAP negative peak occurring during the first three to six responses of each train of stimuli was calculated and in case of significant decremental pattern (CMAP amplitute reduction >10%) correction with edrophonium (2 gg i.p.) was attempted. A group of three mice was injected with each Ig preparation. Limb muscle (deltoid or biceps brachialis) biopsies were obtained from four SNMG patients and routine morphological studies were performed. Statistical analysis was performed by Z2 test.
Seronegative Myasthenia Gravis RESULTS
The series of SNMG patients was made up of 12 males and 26 females (M:F = 1:2.16) with age at disease onset ranging from 6 to 66 yr (mean 32.3). All patients but one had a positive edrophonium test; electrophysiological signs of neuromuscular exaustion (decremental pattern or increased jitter) were present in all cases; electromyographic features of the Lambert-Eaton myasthenic syndrome and repetitive response to single nerve stimuli were absent. In no case was there parents' consanguinity. HLA associations
We did not find any significant difference between patients with seropositive and seronegative generalized MG. The frequencies of B8, DR3 and DQ2 antigens in SNMG were similar to those of controls, while they were significantly higher in SPMG; HLA-DQ3 showed an increased frequency (P<0.01) both in seropositive and in seronegative M G when compared to the general population; HLA-DR2 was weakly associated with SNMG. These data are shown in Table 1. The haplotype B8 DR3 DQ2 was present only in one out of 18 SNMG cases. Clinical data
In 33 of our 38 patients the clinical picture was indistinguishable from SPMG: weakness distribution was typical and daily fluctuations of symptoms were evident. The patients were nine males and 24 females with age at onset ranging from 13 to 66 yr (mean 36.2). Eighteen patients (54%) were affected by mild disease (group 2A according to Osserman), 15 had severe symptoms (groups 2B, 3 and 4). Twentythree patients had undergone thymectomy; the pathological examination of the thymus
157
revealed a normal/involuted thymus in 17 patients (74%), thymic hyperplasia in five (21.7%) and a small nodular thymoma in one (4.3%). An associated autoimmune disease was present in three cases. All patients showed a good response to pyridostigmine, immunosuppressive therapy and plasma-exchange. The efficacy of thymectomy is difficult to assess as it was associated with immunosuppressive therapy in most cases. Plasma IgG from two of these patients transferred the disease to mice (see below). In five patients the clinical picture was characterized by the prevalent involvement of ocular and bulbar districts. All these patients showed opthalmoparesis, marked facial weakness, nasal speech, mild dysphagia and suffered from one or more respiratory crises. Limb muscles were less severely affected, but they also became clearly weakened during exacerbation phases. Clinical course was more stable than in typical MG and, in particular, daily fluctuations were scarcely evident. This group consisted of three males and two females with age at onset ranging from 6 to 23 years (mean 14.8). All patients underwent thymectomy; the histological examination of the thymus showed in no case germinal centres in the medulla. Other autoimmune diseases were not present, but the father of one of these patients was affected by SPMG [17]. The clinical response to pyridostigrnine, although clearly positive in the early stages of the disease, was generally poor with muscarinic side-effects even with small doses; so pyridostigmine was finally withdrawn in three cases. All patients received prednisone with mild benefit and are still under treatment; azathioprine had been administered for at least 1 yr to three patients and then withdrawn as ineffective; thymectomy did not appear to influence the course of the disease. Four patients were submitted to plasma-exchange with deft-
Table 1. HLA associations in seronegative generalized MG, seropositive generalized MG and controls Seronegative MG (n = 18) HLA antigen AI B8 DR2 DR3 DQ2 DQ3
No. 5 2 6 5 6 9*
*Significantly different from controls.
% 27.7 11 33 27.7 33 50
Seropositive MG (n = 49) No. 18 17" 11 23* 26* 18"
Controls (n = 392)
%
No.
%
36.7 34.7 22.4 46.9 53 36.7
102 46 66 77 133 72
26 11.7 16.8 19.6 33.9 18.4
158
A. Evoli et aL
nite improvement, two of them received several plasma-exchange courses which were followed by transient improvement of symptoms. Plasma IgG from two of these patients, when injected into mice, transferred the defect of neuromuscular transmission (see below). Muscle biopsies, from four patients with prevalent oculobulbar symptoms, showed increased fiber diameter variability, scattered focal necrosis and single fiber atrophy mainly affecting type 2; moreover, slight type 1 predominance was present in two of these cases (Fig. 1). In vitro anti-AChR ab production Lymphocyte cultures were performed with PBL from 18 SNMG patients, five of whom were affected by prevalent oculobulbar symptoms. In vitro anti-AChR ab production was never detected in cultures from SNMG patients and controls, while anti-AChR abs were measurable in culture supernatants from five out of 10 SPMG cases. These findings are shown in Fig. 2. Passive transfer studies
Plasma IgG of SNMG patients were from two patients with typical clinical picture and from two with prevalent oculobulbar involvement. RNS induced a significant decremental pattern in mice injected with IgG from SPMG and SNMG patients. The decremental response
occurred after d-tubocurarine injection and was corrected by edrophonium. Figure 3 shows the electrophysiological records in a mouse injected with plasma IgG from a patient with oculobulbar symptoms. No difference was found between mice treated with IgG from SPMG and SNMG patients with respect to the number of mice affected in each group and the severity of decremental response. Neither weakness nor electrophysiological signs of neuromuscular transmission impairment were observed in mice injected with IgG from healthy donors. DISCUSSION
Pathogenetic mechanisms of SNMG are still under debate. Negative results of the anti-AChR assay could be due merely to technical reasons such as too low an ab level or loss of important antigenic determinants in receptor solubilization [6, 9]. Alternatively, the target antigen could be different from AChR. This last hypothesis is supported by those studies reporting lack of correlation between impairment of neuromuscular transmission and end-plate AChR reduction both in intercostal muscle biopsies from SNMG patients and in mice submitted to passive transfer [4, 5]. More recently, Burges et al. have demonstrated that a pre-synaptic defect (reduction in quantal content) is operating in some of these cases [8]. Moreover, the effector mechanisms might be mediated by serum components other than IgG
Fig. 1. (A) HE; (B) ATP-ase pH 4.3. Deltoid biopsy from a SNMG patient with prevalent oculobulbar symptoms. Increased fibre diameter variability and single fibre atrophy (A). Type 1 fibre predominance (B).
Seronegative Myasthenia Gravis anti-AChR ab (fmolll0'
cells)
8-
0
•
•
o o o o
•
•
•
o
o
o
•
•
•
•
•
•
I
nnUili
I
Seronegative MG Seropositive MG
•
•
•
i
t
Controls
Fig. 2. P W M - d r i v e n in vitro synthesis of a n t i - A C h E antibodies.
such as high molecular weight molecules possibly IgM [18], and non-antibody serum factors co-purifying with the IgG fraction [3]. A spontaneous synthesis of anti-AChR ab by PBL cultures implies the presence of auto-reactive B cells. Such a production, under PWM stimulation, has been demonstrated in 50-76% of MG patients [19-21]. In our experience antiAChR abs were undetectable in culture supernatants from SNMG patients. These data are in agreement with previous reports [10, 20] and confirm the results of serum assays. To investigate pathogenetic aspects of SNMG we have compared HLA frequencies in these patients with those in seropositive cases and in the general population. It is well known that, in Caucasians, early-onset non-thymoma MG is strongly associated with HLA-B8, DR3 and DQ2 [19]. In seronegative patients the frequencies of these antigens were not increased, despite the young age at onset. We did not find any significant difference in HLA associations between SPMG and SNMG. HLA-DQ3 occurred at increased frequency in both MG series when compared with the general Italian population. The significance of this association is unknown. A common feature of SNMG is the low frequency of thymic pathology and, in particular, the absence of thymoma [3, 7, 10, 19]. In our experience the association with
159
thymoma, although very rare, can occur; the frequencies of both thymic hyperplasia and involution in our patients are in accordance with other reports. The role of the thymus in SNMG is at present far from clear. Willcox et al. [20] described sparse germinal centers and expanded medullary T-cell areas as distinctive thymus abnormalities in seronegative patients. The authors suggested these changes could be secondary also because thymic cultures from the same patients failed to produce antiAChR antibodies. Thymectomy has generally been reported as scarcely effective in these cases [3, 7, 10, 19, 20] and, consequently, its role in the treatment of SNMG is highly controversial. Clinical heterogeneity of SNMG has been reported by Birmanns et al. [10]: these authors described two forms of the disease and suggested that oculobulbar SNMG which was unresponsive to immunosuppressants could be a non-immune-mediated disorder. Clinical characteristics in our series were similar to those described but all our patients were affected with generalized disease and oculobulbar cases had severe symptoms with respiratory crises. In most of our cases SNMG appears clinically indistinguishable from the seropositive disease; moreover, the efficacy of plasmaexchange and immunosuppressive drugs as well as the results of passive transfer studies leave little doubt that it is an autoimmune disorder. In as much as thymic abnormalities are uncommon in SNMG, the only thymoma and the few thymic hyperplasias were found among these cases. These features suggest that in some patients the disease could be the end of the MG spectrum and that anti-AChE abs could be present at undetectable levels; on the other hand, antigens different from AChR could be involved in other cases. Further clinical experience is necessary to define the therapeutic role of thymectomy. In patients with prevalent oculobulbar involvement the course of the disease is more stable than typical MG and the response to immunosuppressive therapy is less satisfactory. Moreover, thymic abnormalities were absent in every case. In these patients a congenital myasthenia should first be ruled out. The onset of symptoms after the first years of life and the negative family history favor an acquired disease. Muscle biopsy studies showed abnor-
160
A. Evoli et aL
RNS at 3 Hz
To*
TIo
TI7 T3s
TJe
Fig. 3. Decremental response in a mouse injected with plasma IgG from a SNMG patient with prevalent oculobulbar symptoms. *d-Tubocurarine injection; **edrophonium injection. T indicates minutes after injection; calibration bar = 10 inV. Twenty-nine minutes after the injection of d-tubocararine, RNS elicited a significant decremental response which was corrected by the injection of edrophonium.
malities such as focal necrosis and single fiber atrophy which have been described also in muscles from SPMG patients [22]. In two cases these changes were associated with mild type-1 fiber predominance which is a common although not a specific finding in congenital myasthenic syndromes [23]. Immunocytochemical and in vitro electrophysiological studies are necessary to fully investigate the underlying defect of neuromuscular transmission. Nevertheless, the efficacy of plasma-exchange and passive transfer experiments strongly suggest that these cases also are affected by a
humorally mediated disorder. Acknowledgements The financial support from Telethon (grant No. 400) is gratefully acknowledged.
REFERENCES
1. Engel AG. Myasthenia gravis and myasthenic syndromes. Ann Neurol 1984; 16: 519-534. 2. Vincent A, Newsom-Davis J. Acetylcholine receptor antibody as a diagnostic test for myasthenia gravis: results in 153 validated cases and 2967 diagnostic assays. J Neurol Neurosurg Psychiatry 1985; 48: 1246-1252. 3. Verma P K, Oger J J F. Seronegative generalized myasthenia gravis: low frequency of thymic pathology. Neurology 1992; 42: 586-589. 4. Vincent A, Li Z, Hart A, et aL Seronegative myasthenia gravis. Evidence for plasma factor(s) interfering with acetylcholine receptor function. Ann N Y Acad Sci 1993; 681: 529-538. 5. Mossman S, Vincent A, Newsom-Davis J. Myasthenia gravis without acetylcholine receptor antibody: a distinct disease entity. Lancet 1986; i: 116-119. 6. Soliven B C, Lange D J, Penn A S, et al. Seronegative myasthenia gravis. Neurology 1988; 38:514-517. 7. Evoli A, Bartoc~ioni E, Batoc~hi A P, Scuderi F, Tonali P. Anti AChR-negative myasthenia gravis:
Seronegative Myasthenia Gravis
8.
9. 10.
11.
12. 13.
14. 15.
clinical and immunological features. Clin Invest Med 1989; 12: 104-109. BurgesJ, Vincent A, Molenaar P C, Newsom-Davis J, Peer C, Wray D. Passive transfer of seronegative myasthenia gravis to mice. Muscle Nerve 1994; 17: 1393-1400 Drachman D B, De Silva S, Ramsay D, Pestronk A. Humoral pathogenesis of myasthenia gravis. Ann N Y Acad Sci 1987; 505: 90-105. Birmanns B, Brenner T, Abramsky O, Steiner I. Seronegative myasthenia gravis: clinical features, response to therapy and synthesis of acetylcholine receptor antibodies in vitro. J Neurol Sci 1991; 102: 184-189. Lindstrom J. An assay for antibody to human acetylcholine receptor in serum from patients with myasthenia gravis. Clin Immunol Immunopathol 1977; 7: 36-43. Bartoccioni E, Scuderi F, Scoppetta C, et al. Myasthenia gravis, thymectomy and anti-acetylcholine receptor antibody. J Neurol 1980; 224: 9-15. BesingerU A, Toyka K V, Homberg M, Heininger K, Hohlfeld R, Fateh-Moghadam A. Myasthenia gravis: long-term correlation of binding and bungarotoxin blocking antibodies against acetylcholine receptor with changes in disease severity. Neurology 1983; 33:1316-1321. Osserman K E, Genkins G. Studies on myasthenia gravis: review of a twenty-year experience over 1200 patients. Mt Sinai J Med 1971; 38: 497-537. Evoli A, Batocchi A P, Palmisani M T, Lo Monaco M, Tonali P. Long-term results of corticosteroid therapy in myasthenia gravis. Eur Neurol 1992; 32: 37-43.
161
16. Kuks J B M, Limburg P C, Oosterhuis H J G H, The T H. Antibodies to acetylcholine receptor in myasthenia gravis. In vitro synthesis by peripheral blood lymphocytes before and after thymectomy. Clin Exp Immunol 1992; 87: 246-250. 17. Provenzano C, Arancio O, Evoli A, et al. Familial autoimmune myasthenia gravis with different pathogenetic antibodies. J Neurol Neurosurg Psychiat 1988; 51: 1228-1230. 18. Yamamoto T, Vincent A, Ciulla T A, Lang B, Johnston I, Newsom-Davis J. Seronegative myasthenia gravis: a plasma factor inhibiting agonist-induced acetylcholine receptor function copurifies with IgM. Ann Neurol 1991; 30: 550-557. 19. Newsom-Davis J, Willcox N, Schluep M, et al. Immunological heterogeneity and cellular mechanisms in myasthenia gravis. Ann N Y Acad Sci 1987; 505: 12-26. 20. WillcoxN, Schluep M, Ritter M A, Newsom-Davis J. The thymus in seronegative myasthenia gravis. J Neurol 1991; 238: 256-261. 21. Lisak R P, Levinson A I, Zweiman B, Kornstein M J. In vitro synthesis of IgG and antibodies to AChR by peripheral and thymic lymphocytes. Ann N Y Acad Sci 1987; 505: 39-49. 22. Oosterhuis H J G H. Myasthenia gravis. In Clinical Neurology and Neurosurgery Monographs (Vol. 5). Churchill Livingstone, 1984. 23. Engel A G. The investigation of congenital myasthenic syndromes. Ann N Y Acad Sci 1993; 681: 425-434.