Complicating autoimmune diseases in myasthenia gravis: a review

Abstract Myasthenia gravis (MG) is a rare autoimmune disease of skeletal muscle endplates. MG subgroup is relevant for comorbidity, but usually not accounted for. MG patients have an increased risk for complicating autoimmune diseases, most commonly autoimmune thyroid disease, systemic lupus erythematosus and rheumatoid arthritis. In this review, we present concomitant autoimmune disorders associated with the different MG subgroups, and show how this influences treatment and prognosis. Concomitant MG should always be considered in patients with an autoimmune disorder and developing new neuromuscular weakness, fatigue or respiratory failure. When a second autoimmune disorder is suspected, MG should be included as a differential diagnosis.


Introduction
Myasthenia gravis (MG) is a rare autoimmune disease caused by autoantibodies against neuromuscular junction proteins; the nicotinic acetylcholine receptor (AChR), the muscle specific tyrosine kinase (MuSK) or the low-density lipoprotein receptor-related protein 4 (LRP4) [1,2]. MG prevalence of 100-200 per million is similar in most populations [2]. The onset of MG is influenced by age and gender in a bimodal fashion. Females are more commonly affected below the age of 50 years, while males are slightly more frequently affected in late onset MG [3,4]. With improved diagnosis and prolonged survival, the prevalence is increasing, especially in the elderly [5][6][7], also previous under-diagnosed in the elderly. Patients with autoimmune MG can be classified into seven subgroups, each with distinctive autoantibodies and clinical features (Table 1). MG subgroup is highly relevant for comorbidity, but usually not accounted for. MG treatment may influence the risk for other disorders. The course of MG is often complicated by concomitant disease.
Approximately 5% of the population is affected by one or more autoimmune disorders, and the prevalence is higher in females than in males. Patients affected by one autoimmune disorder have a higher risk of developing a second one. MG patients have an increased risk of other autoimmune disorders compared to the non-MG population [8,9]. The frequency of a second autoimmune disorder is 13-22% in MG patients, highest for females and early onset MG (EOMG) [8,9]. Among 100.000 patients with autoimmune disorders, only 35 have MG [10]. However, 0.2% of patients with thyroid disease have MG, much higher than expected by chance [11], whereas 0.1% of patients with multiple sclerosis (MS) have MG [12]. In a systematic review, autoimmune thyroid disease (ATD) was the most frequent of 23 associated autoimmune disorders, occurring in 10% of MG patients [13]. Other common autoimmune associates with MG are systemic lupus erythematosus (SLE) (1-8%), rheumatoid arthritis (RA) (4%), dermato-/polymyositis and Addison's disease [8,[13][14][15][16].
Genetic studies on the development of autoimmune disorders reveal common as well as distinctive susceptibility genes, most strongly at the human leukocyte antigen (HLA) locus [17]. Genome-wide association studies (GWAS) on EOMG shed light on specific additional genetic hot spots for MG [18]. However, the exact pathogenesis, genetic mechanisms and trigger factors underlying MG as well as MG in combination with other autoimmune disorders are still unknown. Patient selection, follow-up time and sensitivity in finding concomitant disease vary in studies of comorbidity. Correct characterization of autoimmune comorbidity will improve insight in MG pathogenesis and optimize the combined treatment of the disorders. and generalized MG subgroups [19], most frequently in the ocular group [20]. In 8% of patients with thyroid-associated ophtalmopathy, AChR antibodies are found [21], possibly due to common autoimmune targets in the eye muscle and/or genetic factors [22,23]. Thymic thyrotropin receptor may act as an auto-antigen in Graves' disease when associated with thymic hyperplasia [24,25]. MG in association with ATD has a younger age of onset, mild clinical expression, rare MG crisis, lower levels of AChR antibodies and lower frequency of thymic changes [20,26,27].
LOMG has a lower frequency of autoimmune overlap than the other MG subgroups [41]. These are the data from a retrospective study of 112 MG patients from 1971 to 2006 conducted in a single neurological department. Nevertheless, both SLE and ATD are more frequent in the LOMG subgroup than in control populations, especially Hashimoto's disease. Their clinical MG features are similar to those of MG patients without autoimmune disorders [28]. Most other autoimmune disorders also occur with a higher frequency in LOMG than in the non-MG population [42]. The prevalence of antiphospholipid antibodies is increased in LOMG, and this should influence decisions regarding thrombosis prophylaxis, for example during MG exacerbations [14].
Myopathy is more common in LOMG than in EOMG, and is associated with anti-titin antibodies [43]. The increased weakness seen in some elderly MG patients can be caused by an immunologically mediated myopathy, leading to a poorer prognosis. In MG patients with inflammatory myopathies, the disease can be severe with bulbar symptoms, MG crisis and often invasive thymoma [44]. There is evidence of CD8+ lymphocyte infiltration of skeletal muscle and presence of anti-striational antibodies [44].

Associated autoimmune disorders in thymoma-MG
The thymus plays an important role for several MG subgroups, and 15% of all MG patients have a thymoma. Patients with MG and thymoma always have anti-AChR antibodies [7]. This reflects that the neoplastic thymus transformation initiates the autoimmune response to AChR [45]. In contrast, the presence of MuSK antibodies in the thymoma-MG subgroup is extremely rare, and it is presumed that thymus is not involved in the pathogenesis of MuSK-MG [46]. Thymomas have a high frequency of autoimmune associated disorders (45%), infections (24%) and both in combination (10%). More than one autoimmune disorders are found in 20% of the cases [47]. However, thymomas are related to only a few autoimmune disorders [48][49][50], the most frequent by far being MG, occurring in 30% of all thymoma cases [47]. In a series of nearly 600 patients with thymoma, 71% had an associated condition, 44% of these had MG, 21% cytopenias, 17% cancer, 6% hypogammaglobulinemia, 5% polymyositis and 2% SLE [51]. In contrast to other MG subgroups, ATD is uncommon in thymoma-MG [52]. Thymoma-MG can be associated also with neuromyotonia, Sjögren's syndrome and autoimmune hemolytic anemia [53].
Thymomas are present at debut of MG. MG can develop many years after the detection of a thymoma. Thymomas can be very large at MG debut, reflecting their presence for many years. MG patients with thymoma are even susceptible to develop autoimmune disorders after thymectomy with removal of the tumor [33,54]. Thymomas seem to have antigen-specific, but also antigen-unspecific, correlation to autoimmunity. T cells of the TCRab lineage occur more frequently in thymomas associated with MG [24,55]. Thymomas produce autoantigen specific T cells by abnormal positive or negative T cell selection, leading to MG, as well as some other autoimmune disorders. These T cells then leave the thymoma to induce autoimmune disease [56]. Risk of thymoma in MG is associated with the TT homozygous genotype and the DNMT3B-579T allele [56]. This genotype does not occur with increased frequency in MG patients without thymoma, even after stratification into MG subgroups [57].
Autoimmune heart disease, especially myocarditis and giant cell myocarditis, has a higher incidence in the thymoma-MG subgroup, confirmed pathologically with lymphocyte infiltration and giant cells in the heart [58]. Heart muscle antibodies are related only to thymoma-MG and LOMG subgroups [59]. Antibodies such as those against voltagegated potassium channel (VGKC) Kv1.4, titin and ryanodine receptor have been shown to react with heart muscle in MG patients [58]. A possible association between anti-Kv1.4 antibodies and cardiac involvement, especially myocarditis, was recently described [60], the disease characterized by heart failure and arrhythmias together with limb and paraspinal muscle weakness. Giant cell myocarditis is potentially fatal and should be paid special attention [61], with immediate admission to an intensive care unit and start of immunosuppressive treatment [62].
Neoplasia, hyperplasia and age-related degeneration of the thymus are relevant when considering MG treatment. In some patients, immunosuppressive MG treatment, including thymectomy, may disturb the balance of immunity and lead to additional disorders [11,54,63].
The T cell repertoire is altered after thymectomy [33]. Theoretically this could both increase and decrease the risk for a second autoimmune disorder. The MG response to thymectomy is the same with and without a second autoimmune disease. However, thymectomy has no influence on the second autoimmune disease [11,64]. Studies suggesting that thymectomy may increase the susceptibility for an additional autoimmune disorder, and especially for SLE [65,66], are based on a small number of cases and with a selection bias. MG appears in about 1% of the patients post thymomectomy, and surgery may not prevent postoperative MG [67].

Associated autoimmune disorders in MuSK-MG
MuSK antibodies are present in 10-70% of all MG patients without AChR antibodies [2,68,69], and is clinically more severe than other MG subtypes with involvement of facial, bulbar and upper body muscles, sometimes with muscle atrophy [2,69,70]. MuSK-MG is more prevalent in Mediterranean countries and less in Northern Europe [71].
Very few studies have reported the frequency of associated autoimmune disorders in MuSK-MG specifically. Association with HLA-DR14-DQ5 occurs for MuSK-MG, similar to pemphigus foliaceus and pemphigus vulgaris [72][73][74]. Interleukins 4 and 13, secreted by Th2-type T cells, are involved in MuSK-MG and pemphigus [75]. MuSK-MG and SLE can be found in the same patient [19]. An association between relapsing-remitting MS and MuSK-MG has been suggested [76]. Rituximab represents an effective treatment [77].
Coexistence of AChR and MuSK antibodies has been reported in only five patients [78][79][80][81][82]. In one case, AChR, MuSK and VGKC antibodies co-occurred with MG and Morvan's syndrome without any thymoma [78]. In the other four cases, the patients were AChR antibody positive at disease onset, but became MuSK antibody positive after thymectomy. D-penicillamine can induce both AChR and MuSK antibody-positive MG, a rare phenomenon which is reversible after discontinuation of the treatment [83].

Associated autoimmune disorders in LRP4-MG
LRP4 is a receptor for agrin and important in the formation of the neuromuscular junction [1]. LRP4 autoantibodies are reported in up to 50% of MG patients without AChR or MuSK antibodies, and even in a few patients with MuSK-MG [1,84,85]. Patients with exclusively bulbar symptoms may have double seropositive MG (LRP4 and AChR) [86]. Case reports of concomitant autoimmune disease in the LRP4-MG subgroup are still lacking. However, LRP4 antibodies have been found in Lambert Eaton myasthenic syndrome and neuromyelitis optica [68]. Thymoma has not been observed in LRP4-MG [85]. As for MuSK-MG, no role of the thymus has been established in the LRP4-MG pathogenesis [46,87]. The IgG1 subclass, a complement activator, predominates for the LRP4 antibodies. In AChR-MG, the IgG1 and IgG3 subclasses predominate. In contrast, MuSK-MG predominantly has IgG4 which does not activate complement [1,68]. The clinical presentation of LRP4-MG is milder compared with other MG subgroups, and the response to therapy is good. A significant number of double positives (7.5% AChR/ LRP4-MG and 15% MuSK/LRP4-MG) were recently described, this being more severe cases [88].

Autoimmune overlap and common etiologic factors
Autoimmunity arises from an inappropriate response against tissue elements normally present, leading to complex chronic diseases characterized by the loss of immune tolerance to self-antigens [89]. Autoimmune overlap reflects common pathogenic mechanisms [10,90]; immunological factors leading to activation of autoreactive B and T cells, genetic susceptibility involving specific candidate genes and epigenetic factors [91]. The genetic interplay is highly recognized in the development of autoimmune disorders, due to high concordance rates reported from twin studies and observations of candidate genes, particularly HLA genes, strongly associated with autoimmune diseases [91]. Understanding the common mechanisms in the pathophysiology of MG and other autoimmune disorders is important for an effective and target-directed treatment [92]. Associated genetic variants in autoimmune disease are shared between systemic and organspecific autoimmune diseases [93]. Disease models for autoimmune diseases such as MG, RA, MS, DM, autoimmune myocarditis, ATD and uveitis have been developed using HLA alleles transgenic mice [94]. This humanized transgenic mice models help in understanding the role of HLA genes in the pathogenesis of MG and other autoimmune diseases.
The genetic influence of the major histocompatibility complex (MHC) in MG is thought to be mediated through a single signal in the class I region, induced in the vicinity of the HLA complex protein 5 gene located between MICA and MICB [95,96]. EOMG and other autoimmune disorders are associated with the HLA A1-B8-DR3-DQ2 haplotype. GWAS analysis using 649 EOMG cases and 2596 controls showed a strong association to the HLA-B * 08 locus allele and to two additional loci outside MHC: the PTPN22 gene and TNFAIP3-interacting protein 1 (TNIP1) [18]. The HLA-B * 08 represents a risk allele for most autoimmune diseases [97]. In vivo and in vitro studies suggest that TNIP1 plays a crucial role in regulating several cellular pathways involved in immune-related disorders [92]. TNIP1 is a risk allele for RA, psoriasis, Sjogren's syndrome and SLE [98][99][100].
It will be of interest to determine whether endophenotypes, such as the identity of the autoantibody (AChR vs. MuSK vs. titin), the presence of thymoma or the occurrence of associated autoimmune diseases are associated with distinctive HLA-region signals. MHC class II gene associations have been reported specifically for patients with anti-MuSK or anti-titin antibodies [103]. The HLA class II allele HLA-DRB1*16,-DRB1*14 and -DQB1*05 allele is linked to MuSK-MG [73,103,104], and also HLA-DRB1*03 appears to have a distinguishing role for this subgroup compared to AChR-MG [103]. Genetic susceptibility and MG subgroups vary between populations. More than 90% of Southern Han Chinese ocular MG patients had the DQ9 haplotype [105]. Some non-major MHC genes identified as important susceptibility genes in MG are shared with other autoimmune disorders. The protein tyrosine phosphatase nonreceptor 22 (PTPN22) R620W gene polymorphism is a general risk factor for autoimmune diseases with an increased production of auto-antibodies [18] and predisposes MG and EOMG in particular for additional autoimmune diseases [29,106]. PTPN22 showed the strongest association to the thymoma-MG and EOMG subgroups [107].
The lack of association of PTPN22 R620W polymorphism with MuSK-MG and antibody negative-MG subgroups [108] emphasizes the different genetic background for MG subgroups. Young EOMG (debut520 years) is strongly associated with the nicotinic cholinergic receptor alpha 1 locus which encodes the a-subunit of the AChR and interacts with the autoimmune regulator [109]. CTLA4 gene polymorphisms are associated with MG and also other autoimmune diseases such as type 1 DM, ATD, SLE, RA and celiac disease [92,[110][111][112][113][114][115]. The cathepsin L2 is associated with the EOMG subgroup and type 1 DM [36].

Conclusions
Autoimmune overlap between MG and other autoimmune disorders reflects common pathogenic mechanisms. ATD is the most frequent comorbidity. Comorbidity is linked to MG subgroup. It is a paradox that LOMG and thymoma-MG have a broad antimuscle antibody response, but the immune response usually being confined to muscle, whereas EOMG has a restricted antimuscle response against AChR only, but often has a broad general response against several organ tissues. MG in association with other autoimmune disorders has a less favorable prognosis. It is important to consider coexisting MG in patients with autoimmune disorders and neuromuscular weakness, fatigue and respiratory failure.

Declaration of interest
This review was supported by European Federation of Neurological Societies Fellowship Programme. None declared.