Clinical Features and Diagnostic Usefulness of Antibodies to Clustered Acetylcholine Receptors in the Diagnosis of Seronegative Myasthenia Gravis

Importance  Cell-based assays (CBAs) were shown to improve detection of acetylcholine receptor (AChR) antibodies in patients with myasthenia gravis (MG). Herein, we asked whether these assays were able to help determine the diagnosis in patients studied in routine clinical practice.

Objectives  To determine the diagnostic usefulness of CBAs in the diagnosis of MG and to compare the clinical features of patients with antibodies only to clustered AChRs with those of patients with seronegative MG (SNMG).

Design, Setting, and Participants  All patients with clinical suspicion of MG who were seen within the Division of Clinical Neurology at the John Radcliffe Hospital in Oxford, England, between November 1, 2009, and November 30, 2013. Their serum antibodies and clinical features were studied.

Exposures  Radioimmunoprecipitation assay (RIPA) and CBA were used to test for standard AChR antibodies and antibodies to clustered AChRs in 138 patients. All available samples from patients with SNMG were retrospectively tested for lipoprotein receptor–related protein 4 (LRP4) antibodies.

Main Outcomes and Measures  Demographic, clinical, neurophysiological, and laboratory data.

Results  In total, 138 patients were tested for antibodies to clustered AChRs, and 42 had a final diagnosis of MG. The clustered AChR CBA detected antibodies in 38.1% (16 of 42) of RIPA-negative patients with MG with 100% specificity. All patients with SNMG who were tested for LRP4 antibodies (21 of 26) were negative by CBA. Compared with patients with SNMG, patients with antibodies only to clustered AChRs had frequent prepubertal onset (62.5% [median age, 6 years; age range, 1-52 years] vs 11.5% [median age, 38 years; age range, 2-72 years], P ≤ .05), high prevalence of ocular MG (62.5% vs 42.3%), milder disease severity with less bulbar involvement (25.0% vs 46.2%), and absence of respiratory symptoms (0% vs 23.1%). Response to treatment and prognosis was good, with a reduced need for thymectomy (6.3% vs 19.2%) and a high proportion of patients going into remission (50.0% vs 8.3%, P ≤ .05). These observations also apply to the classic AChR MG phenotype seen in large series.

Conclusions and Relevance  Cell-based assay is a useful procedure in the routine diagnosis of RIPA-negative MG, particularly in children. Patients with antibodies only to clustered AChRs appear to be younger and have milder disease than other patients with MG. These observations will have implications in planning treatment.

Quiz Ref IDMyasthenia gravis (MG) is an antibody-mediated autoimmune disease of the neuromuscular junction. Approximately 80% of patients with generalized MG have autoantibodies against the muscle nicotinic acetylcholine receptor (AChR) measured by radioimmunoprecipitation assay (RIPA).^1,2 The AChR antibodies are predominantly IgG1 and IgG3 subclasses and produce severe loss of AChRs by complement-mediated damage to the postsynaptic membrane, receptor endocytosis, and occasionally direct AChR block.³ These patients are classically referred to as having seropositive MG. Quiz Ref IDOther patients with generalized MG and 50% of those with ocular MG lack detectable AChR antibodies by RIPA.⁴

Autoantibodies to muscle-specific tyrosine kinase (MuSK) measured by RIPA have been reported in a variable proportion of patients with seronegative MG (SNMG), ranging from 0 to 70%.^5-12 Autoantibodies to MuSK are mainly of the IgG4 subclass, which does not activate complement. They prevent the interaction of MuSK with lipoprotein receptor–related protein 4 (LRP4) and therefore inhibit agrin-dependent AChR clustering.¹³ MuSK autoantibodies identified patients with distinctive clinical features.^14,15Quiz Ref ID Patients with MG lacking detectable AChR and MuSK antibodies by RIPA are referred to as having SNMG.

A cell-based assay (CBA) was established for the improved detection of AChR antibodies in patients previously seronegative by RIPA.¹⁶ Cell-based assays also measure AChR antibodies detected by RIPA,¹⁶ although this is not performed routinely because of the high cost and time-consuming nature compared with the RIPA. The CBA involves expressing AChRs on the surface of a human embryonic kidney (HEK) cell and clustering by coexpression with the intracellular anchoring protein rapsyn. This is performed by transfecting the live HEK cells with the appropriate complementary DNAs (cDNAs) encoding these proteins. The binding of AChR antibodies can be scored visually using indirect immunofluorescence. Unlike most other diagnostic antibody tests, this CBA allows detection of antibodies binding to AChRs in a natural membrane environment, where they adopt native conformational states and appropriate glycosylation levels and are clustered as they are at the neuromuscular junction. The proportion of patients with SNMG with autoantibodies to clustered AChRs ranges from 16% to 60%.^16-18 These antibodies are mainly of the complement-fixing IgG1 subtype and have pathogenic mechanisms similar to those detected by RIPA.¹⁷ Subsequently, several groups have reported autoantibodies against other components of the neuromuscular junction, namely, agrin, LRP4, and collagen Q in a variable and generally low proportion of patients with SNMG.^19-23

We aimed to assess the clinical usefulness of clustered AChR antibodies in MG. We also sought to describe the clinical features of patients seen within the Division of Clinical Neurology at the John Radcliffe Hospital in Oxford, England, who had been tested for clustered AChR antibodies since the implementation of the assay.

We assessed all 138 patients seen within the Division of Clinical Neurology at the John Radcliffe Hospital who were tested for antibodies against clustered AChRs between November 2009 and November 2013, all of whom had been negative by RIPA for AChR and MuSK antibodies (Figure 1). Myasthenia gravis was diagnosed by clinical and electromyographic criteria. In patients with normal neurophysiology and lacking detectable autoantibodies, MG was diagnosed based on the presence of fatigable weakness and a positive response to treatment with cholinesterase inhibitors or immunosuppression by experienced myasthenia physicians (S.J., C.B., D.H.-J., M.I.L., and J.P.). This study was approved by the Clinical Audit Team at the John Radcliffe Hospital.

Patients with MG lacking detectable AChR antibodies by RIPA or CBA or lacking MuSK antibodies by RIPA were referred to as having SNMG. All available samples from these patients were retrospectively tested for LRP4 antibodies using a CBA. The patients in whom a final diagnosis was not achieved were referred to as having an uncertain diagnosis. If a final diagnosis other than MG was made, the patients were referred to as having other diagnoses. The clinical features of patients with SNMG were compared with those of patients with MG with antibodies only to clustered AChRs. Data on the individual patients with MG, including demographic data, were collated. Race/ethnicity is self-assessed on a routine basis in all our patients as part of a general questionnaire using options defined by the investigators and was reported because of the unexpected high proportion of nonwhite patients. Clinical features were graded on a scale of 0 to 3 (0 is absent, 1 is mild, 2 is moderate, and 3 is severe). The distribution and severity of myasthenic weakness were classified according to the Myasthenia Gravis Foundation of America (MGFA) grading system.²⁴ Clinical state of patients with MG after institution oftreatment was classified according to the MGFA postintervention status.²⁴

For the clustered AChR assay, HEK cells were transfected with cDNAs expressing human AChR α, β, δ, and ε/γ subunits and rapsyn-enhanced green fluorescent protein in a ratio 2:1:1:1:1. For the LRP4 assay, HEK cells were transfected with cDNAs expressing human LRP4 and a chaperone protein (low-density lipoprotein receptor–related protein–associated protein 1) to enhance cell surface expression. Measurement of antibody binding was performed by indirect immunofluorescence as previously described.¹⁶ Results were measured by 2 observers (S.H., L.J., M.W., and M.I.L.) on a nonlinear visual scale from 0 to 4 (0 is no signal, 0.5 is unclear, 1 is weak positive, 2 is moderate positive, 3 is strong positive, and 4 is very strong positive). When the score differed, the mean result was given. Unclear results were repeated until consensus was achieved.

In total, 138 patients were tested for clustered AChR antibodies during the period (Figure 1). Forty-two (30.4%) had a final clinical diagnosis of MG, 51 (37.0%) had an uncertain diagnosis, and 45 (32.6%) had other diagnoses (Table 1). The patients with a clinical diagnosis of MG represented all the AChR antibody–negative and MuSK-negative MG cases by RIPA encountered during the study period.

All patients with uncertain diagnosis and other diagnoses were negative for clustered AChR antibodies. In the MG group, 16 of 42 (38.1%) were positive for clustered AChR antibodies (Figure 1). One patient with an unclear CBA result was designated as having SNMG. All patients with SNMG who were tested for LRP4 antibodies (21 of 26) were negative by CBA.

The clinical characteristics of 16 patients with clustered AChR antibodies (CBA score range, 1-3; median score, 2) are listed in Table 2 and Table 3. Ten of 16 were female. The age range was wide (age range, 1-52 years), and 62.5% were children (age range, 1-10 years), with 8 of 10 children initially seen before age 5 years. Seven patients (43.8%) were of black British, Caribbean, or African race/ethnicity (4 individuals) or mixed white Asian (3 individuals). The rest were white British or white other. Neurophysiology at diagnosis was abnormal in 8 patients (50.0%), normal in 6 patients (37.5%), and not assessed in 2 patients (12.5%). Chest imaging was normal in the 3 patients who were imaged.

The presentations were predominantly ocular (62.5%) with ptosis and restricted eye movements, with no generalization during the follow-up period. Bulbar symptoms were scarce and mild (25.0%), and no patients had respiratory weakness. The maximum MGFA grades related to age at onset are shown in Figure 2A.

Most of the cases were successfully treated with pyridostigmine bromide and prednisolone or with pyridostigmine alone, and only 2 of 16 patients (12.5%) needed further immunosuppression with azathioprine. Five patients were receiving immunosuppressive treatment at the time of antibody testing. Thymectomy was performed only in 1 patient, who had the highest MGFA grade (IVA). Pathological examination showed an atrophic thymus, with no evidence of thymoma or lymphocytic infiltrations.

The mean (SD) follow-up period was 54.5 (40.6) months. Among 14 patients with adequate follow-up, the outcome was good in all but 3 patients, with complete stable remission in 4, pharmacological remission in 3, and minimal manifestations in 4. The mean (SD) time from onset of symptoms to complete resolution was 78.75 (55.33) months. However, 3 children with ocular or mild generalized disease did not attain remission.

The clinical characteristics of 26 patients with SNMG with negative CBA results are summarized in Table 4 and Figure 2B and C. Overall, the patients with clustered AChR antibodies had younger age at onset (P = .02, Mann-Whitney test) and a trend toward milder disease (P = .06, Fisher exact test), and a higher proportion attained clinical remission (P = .03, Fisher exact test) compared with the patients with SNMG who did not have clustered AChR antibodies.

We confirm that the clustered AChR CBA improves diagnostic sensitivity for MG. We show that patients with these antibodies have a milder phenotype than patients without any detectable AChR antibodies, with younger onset age and complete or pharmacological remission in 7 of 14 patients with adequate follow-up.

Quiz Ref IDCell-based assays were recently established for the improved detection of AChR antibodies in patients who were previously seronegative. Herein, we show that the CBA is helpful in the diagnostic workup of MG, detecting AChR antibodies in 38.1% (16 of 42) of negative patients with MG with 100% specificity. On the other hand, none of the patients with SNMG who were tested for LRP4 antibodies were positive. This is in keeping with our ongoing observations that the incidence of this antibody is low.

Autoantibodies detected by CBA have pathogenic mechanisms similar to those detected by RIPA and include complement-mediated lysis.^16,17 Complement deposition was noted in a passive transfer model of clustered AChR antibody MG.¹⁷ Although not formally described, it is likely that the IgG1 antibodies detected by CBA cross-link the clustered AChRs on the cell surface by divalent binding, thus explaining the increased sensitivity of the CBA. Alternatively, a less likely explanation could be the loss of antigenic determinants in the solubilized AChRs used in the RIPA. Cell-based assays require the use of live cells, tissue culture facilities, and expertise with performing and interpreting the assay, which are scored visually using indirect immunofluorescence. Because of these limitations, their current use is mainly confined to specialized research centers. However, if these difficulties were overcome, fluorescence-activated cell sorting analysis¹⁶ could be a suitable option for an automated and objective analysis.

Patients with antibodies only to clustered AChRs had early onset and overall milder disease, with predominantly isolated ocular symptoms, low generalization rate, better prognosis, and no need for thymectomy. These results differ from those of patients with SNMG in the present study and differ from the classic AChR MG phenotype seen in large series with respect to age at onset, MGFA grades at maximum severity, and prognosis.^25,26 However, because of the small numbers of patients included herein, their low MGFA grades were not significantly different compared with patients with SNMG (P = .06). Our group previously found that 50% of patients with RIPA-seronegative ocular MG had complement-fixing IgG clustered AChR antibodies.¹⁷ Results from 2 other studies^18,27 have suggested that individuals with antibodies to clustered AChRs have a milder disease, but those studies did not include pediatric populations.

Juvenile MG, especially in prepubertal children, is reported to differ from adult MG, with more ocular myasthenia, lower rates of generalization, better prognosis with a higher probability of achieving remission, and a lower frequency of AChR antibodies by RIPA.^28-31 However, the previous observation of similar female and male prevalence^28,29 differs from our 2:1 ratio in those with clustered AChR positivity. Seronegative MG was more common in prepubertal (44%) compared with peripubertal (18%) and postpubertal (0%) onset of MG.³² Although it is possible that some of those individuals had an undiagnosed congenital myasthenic syndrome, we also found a high proportion of prepubertal onset of MG (66.6%) in the group of patients with autoantibodies against clustered AChRs who were previously negative on RIPA. Therefore, the usefulness of the CBA is shown with respect to the differential diagnosis with congenital myasthenic syndrome because this is the major group with an alternative diagnosis and both are rare disorders in childhood.³³

Thymectomy should be considered as a treatment option early in the course of generalized AChR antibody–positive juvenile MG.³⁴ However, indications for thymectomy within the prepubertal and seronegative groups are less clear.³⁵ In our cohort of pediatric patients with antibodies only to clustered AChRs, thymectomy was not performed, probably related to the mildness of disease, high rate of pharmacological remission, and young age of some patients. In the French study¹⁸ cited earlier, thymectomy was performed in 2 individuals (ages 14 and 18 years) with antibodies to clustered AChRs and revealed thymus hyperplasia and thymoma. A further study¹⁶ in adult patients reported evidence of thymic involvement, with typical lymphocytic infiltrates and germinal centers less striking but in a fashion similar to the classic AChR MG.

The reasons underlying the clinical differences between prepubertal juvenile MG and adult MG are unknown. We found a female predominance and a high proportion of nonwhite patients (43.7%) within our cohort, especially black British individuals. This agrees with the observation that juvenile MG is rare in patients of white race/ethnicity.³³ In Asian populations, juvenile MG and ocular MG have been reported to represent almost half of the MG cases.³⁶ Chinese racial/ethnic groups have also been reported to have lower antibody titers on RIPA, which could be related to a higher frequency of ocular cases or milder disease.³⁷

Quiz Ref IDIn conclusion, we have shown that measuring antibodies to clustered AChRs by CBA is helpful for the diagnosis of SNMG. This is particularly true in children, who tend to have early onset, milder disease, and increased chance of spontaneous remission.

Accepted for Publication: February 17, 2015.

Corresponding Author: Jacqueline Palace, DM, Nuffield Department of Clinical Neurosciences, University of Oxford, John Radcliffe Hospital, Level 3, West Wing, Headley Way, Headington OX3 9DU, England (jacqueline.palace@ndcn.ox.ac.uk).

Published Online: April 20, 2015. doi:10.1001/jamaneurol.2015.0203.

Author Contributions: Drs Vincent and Palace had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Drs Vincent, Leite, and Palace contributed equally to this work.

Study concept and design: Rodríguez Cruz, Al-Hajjar, Beeson, Vincent, Leite, Palace.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Rodríguez Cruz, Al-Hajjar, Vincent, Palace.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Rodríguez Cruz.

Administrative, technical, or material support: Huda, Jacobson, Woodhall, Beeson.

Study supervision: Vincent, Palace.

Conflict of Interest Disclosures: Dr Rodríguez Cruz reported being supported by the National Health Service National Specialised Commissioning Group for Congenital Myasthenia. Dr Huda reported being supported by a Watney, Myasthenia Gravis Association, and Oxford Biomedical Research Centre fellowship. Ms Buckley reported receiving research support from the Medical Research Council, United Kingdom. Dr Vincent reported serving on scientific advisory boards for the Patrick Berthoud Charitable Trust, the Brain Research Trust, and the Myasthenia Gravis Foundation of America; reported receiving funding for travel and a speaker honorarium from Baxter International, Inc, and Biogen, Idec; reported being an associate editor for Brain; reported earning royalties from the publication of Clinical Neuroimmunology (Blackwell Publishing, 2005) and Inflammatory and Autoimmune Disorders of the Nervous System in Children (Mac Keith Press, 2010); reported obtaining research support from the European Union, National Institute for Health Research Oxford Biomedical Research Centre, Euroimmun AG, and the Sir Halley Stewart Trust; and reported receiving MuSK antibody royalties and consulting fees from Athena Diagnostics, Inc. Dr Leite reported being supported by the National Health Service National Specialised Commissioning Group for Neuromyelitis Optica and by the National Institute for Health Research Oxford Biomedical Research Centre and reported receiving speaking honoraria from Biogen Idec and travel grants from Novartis. Dr Palace reported being partly funded by highly specialized services to run a national congenital myasthenia service and a neuromyelitis optica service; reported receiving support for scientific meetings and honoraria for advisory work from Merck Serono, Biogen Idec, Novartis, Teva, Chugai Pharma, and Bayer Schering and unrestricted grants from Merck Serono, Novartis, Biogen Idec, and Bayer Schering; reported that her hospital trust receives funds for her role as a clinical lead for the United Kingdom Department of Health risk-sharing scheme; reported receiving grants from the National Multiple Sclerosis Society and The Guthie-Jackson Charitable Foundation for unrelated research studies; and reported serving as a board member for the charitable European MS foundation “The Charcot Foundation” and on the steering committee for the European collaborative multiple sclerosis imaging group “Magnetic Resonance Imaging in Multiple Sclerosis.” No other disclosures were reported.