The epidemiology of myasthenia gravis

Neuromuscular junction (NMJ) disorders include several dysfunctions that ultimately lead to muscle weakness. Myasthenia gravis (MG) is the most prevalent NMJ disorder with a highly polymorphic clinical presentation and many different faces. Being an autoimmune disease, MG correlates with the presence of detectable antibodies directed against the acetylcholine receptor, muscle-specific kinase, lipoprotein-related protein 4, agrin, titin, and ryanodine in the postsynaptic membrane at the NMJ. MG has become a prototype serving to understand both autoimmunity and the function of the NMJ better. The aim of this review is to synthesize some of the epidemiological data available. Epidemiological data regarding MG are important for postulating hypotheses regarding its etiology and facilitating the description of MG subtypes. Thus, adequate documentation through broad databases is essential. The incidence and prevalence of MG reported around the globe have been rising steadily and consistently over the past decades. Ethnic aspects, gender-related differences, and environmental risk factors have been described, implying that these might contribute to a specific phenotype, further suggesting that MG may be considered an umbrella term that covers several clinical entities.


INTRODUCTION
Neuromuscular junction (NMJ) disorders comprise several dysfunctions that ultimately lead to muscle weakness. Some of these diseases, such as congenital myasthenic syndromes, are genetic. Toxins like botulinum toxin and curare also disrupt the activity at the level of the NMJ. Other NMJ disorders are acquired autoimmune forms, such as myasthenia gravis (MG), Lambert-Eaton myasthenic syndrome (LEMS), and neuromyotonia.
MG is the most frequent NMJ disorder with a highly polymorphic clinical presentation manifesting predominantly, as the etymology suggests ("mus" meaning "muscle", "asthenia" meaning "weakness"), with muscle weakness and fatigue [1]. MG represents the largest group of NMJ disorders and occurs as a result of the impairment of neuromuscular transmission [2]. Given its pathogenetic mechanism and specific anatomical location, MG has become both a prototype for autoimmune disorders and a model for understanding the synaptic function [3]. It is an autoimmune B-cell mediated disease and correlates with the presence of antibodies directed against the acetylcholine receptor (AChR), muscle-specific kinase (MuSK), lipoprotein-related protein 4 (LRP4), agrin, titin, and ryanodine in the postsynaptic membrane at the NMJ [3]. MG patients may be categorized into subtypes in order to facilitate the therapeutic process

Incidence and prevalence
Based on 35 studies up to 2007 [1], the incidence rate of MG varied from 1.7 to 21.3, with a global rate of 5.3 per million person-years. The pooled incidence rate after 1976 is approximately twice greater than the one before 1976, which is 6.5 vs. 3.5, respectively [7]. In Table 1, we updated this review to 2019, adding 29 studies with a range of 0.15 to 61.33 per million person-years. The global incidence rate of acetylcholine receptor antibody-positive MG ranges between 4 and 18 per million person-years [8]. The incidence of MuSK MG is estimated at 0.1 per million person-years in Holland and 0.32 per million person-years in Greece [1].
Data on MG incidence have varied over time and among different geographical regions, questioning whether there is a real geographical variation (that could point to the disease's etiology) or if this is due to methodological biases. Most epidemiological studies regarding rare and heterogeneous diseases such as MG bear limitations like small study populations, different inclusion criteria and sources of data, disparate diagnosis criteria, and often provide data that cannot be compared. Nationwide databases, including whole populations, offer a reliable ground for epidemiological studies but are not available in most countries.
The discrepancy in incidence due to methodological biases is expected to disappear as studies expand their year span and their quality enhances, thus possibly revealing real geographical trends.
Depending on the geographic location, the prevalence of MG ranges between 1.5 to 17.9 [1], or between and 2.19 to 36.71 cases/100.000 population (Table 1). This indicates an estimate of 56,000-123,000 patients in Europe [8] and 60.000 in the United States [5].
No epidemiological study has been performed in the Republic of Moldova so far; the only data available come from the neuromuscular division of the Institute of Neurology and Neurosurgery in Chisinau, where 320 patients are being supervised through an MG database, which includes patients diagnosed since 2008. One study determining the prevalence of multiple sclerosis in the Republic of Moldova has been performed [37], laying the ground for further similar epidemiological studies.
Epidemiological studies of MG in Kаzakhstan have not been published yet. According to preliminary data, the prevalence ratio of MG in Kazakhstan is estimated to range from 0.5 to 5.0 per 100.000 people [38]. MG patients are being supervised through the Electronic Register of Dispensary Patients, which included 1187 registered patients in July 2019. In Almaty, the number of patients aged 18 years and older increased by 24% between 2017 and 2019. The largest number of patients was registered in Almaty (112 people), followed by the East Kazakhstan region -(103 patients) [38]. The growing number of cases and lack of comprehensive information on MG in Kazakhstan prompt to study it further to raise the effectiveness of diagnosis and treatment of this rare pathology and assess its social and economic implications.
Before 1934, the prevalence of MG was estimated at 1 in 200.000. After 1934, with the introduction of anticholinesterase compounds, it rose to 1 per 20.000 and 1 per 17.000 population after detecting serum antibodies against AChR in 1969 [39].
In the last decades, a further steady increase in prevalence has been observed [27]. The evidence comes from several geographic areas and diverse ethnicities [40]. Prolonged survival, a normal life expectancy, an aging population as well as improved diagnosis are probably the main explanations [9,24]. Clinicians may be noticing that the natural history of the disease is different in older patients, questioning again whether MG is an umbrella term for several distinct diseases [5].
Currently, mortality from the disease is 5-9% [39], the overall in-hospital mortality rate is 2.2% and 4.7% in a myasthenic crisis. The most important predictors of death are age and respiratory failure [41]. The mortality rate is slightly higher in males (14%) than females (11%) [39]. The underlying or contributing cause of death among patients with MG are respiratory tract diseases, such as pneumonia and influenza. The occurrence of cardiac, cerebrovascular, and malignant diseases seems to be lower in MG patients [42]. Compared to 1934, when the few patients diagnosed with MG were those with extreme weakness and 70% of them died of respiratory failure or pneumonia within two years [39], the term "gravis" is now considered obsolete.
Overall, we expect to see more MG patients in the future [40]. This tendency implies that the future MG patient will be older and present with more comorbidities, which will lead to a more elaborate treatment.

Ethnicity
A certain racial or ethnic background may influence MG's presentation and course. For example, the age of onset is higher in Caucasians than non-Caucasians [43]. Also, the female gender is more prevalent among Hispanic, Asian, and African-American ethnicities compared to Caucasians [43].
The incidence rate of MG was reported to be higher in African-American women (0.01 per 1.000 population/year) compared to Caucasian women and Caucasian and African-American men (0.007-0.009 per 1.000 population/year) [41], probably because of a predisposition in the Black population for developing autoimmune diseases [43]. However, this study only dealt with hospitalized MG patients. There is also a higher incidence of ocular MG among Black men and women [44], while Caucasian women show higher rates of generalized MG [43].
A multi-racial study has been conducted in South Africa [45], showing that Black subjects of indigenous African origin were more likely than Whites to develop ocular MG. White subjects were more likely than Black subjects to develop severe generalized MG that was poorly responsive to treatment and repeated myasthenic crises [45].
Asian populations report more cases of early-onset MG, which is more often ocular MG [3]. Juvenile-onset and particularly infantile-onset ocular myasthenia (0-4 years) [23] is also more common in Asian populations, with up to 30% of patients manifesting this form [3]. This epidemiological particularity has been associated with HLA-Bw46 and DR9 [3]. Only 50% of juvenile-onset MG in Asian populations are anti-AchR positive [46].
In a Japanese epidemiological study [23], 80.6% of children in the infantile-onset group presented with the ocular form, compared to 14-30% in Europe and North America. A similar high frequency of juvenile-onset ocular MG was reported in Chinese populations, suggesting a distinct immunological mechanism [23].
The late-onset peak of MG is less expressed in Japanese, Indian and Chinese patients with MG [3]. A high proportion of patients from Asia presents higher rates of MuSK MG compared with Caucasian MG patients [47].
Ethnic epidemiological particularities were also studied in a cohort of Jewish Ashkenazi (ASH) and non-Ashkenazi (NASH) because of their different origins [47]. ASH Jews originated in Europe, while NASH Jews originated in the Middle East and North Africa, having a more diverse ancestry. The results revealed a higher frequency of ocular MG in the ASH group [47].
Studies of migrating populations are useful for exploring the genetic and environmental aspects of multifactorial diseases. A study performed in Norway and Netherlands tried to determine the epidemiological characteristics of MG in immigrants compared to native patients [26]. The study showed that MuSK MG and MG with thymoma were more frequent in Asian immigrants in comparison to other racial types indicating that these patients bring genetic elements or lifestyle particularities acquired before the migration that support their distinctive phenotype.

Age and gender
MG affects all ages, but it is considered "a disease of young women and old men". The earliest age of onset reported is 1 year [48], while the oldest onset age reported was 98 years old [3]. There has been a constant increase of late-onset MG in Western [23] and Asian [49] countries over the last 30 years. The incidence of MG patients with an onset age of 50 years or more increased by 1.5-fold, while the number of patients with an onset age of 65 years or more showed a 2.3-fold rise [23]. This increase is probably due to better recognition of the disease and improved diagnostic tests. However, it may also suggest a distinct immunological background in the elderly or that environmental factors may be involved.
The most common onset age is between 20 and 39 years in women [50] and between 50 and 70 years in men [3]. The onset age curve is constantly bimodal for females, with an early-onset and a late-onset peak, and tends to present only one late-onset peak for males. The cause of this distribution has haunted researchers for many years. Gender is a deciding epidemiological risk factor for the development of autoimmune diseases, given that the immune system is a subject of sexual dimorphism [51]. Many autoimmune diseases, including MG, are more prevalent in women than in men. In early-onset MG, the female:male ratio is 3:1 [52]. These observations raise the question of estrogens as mediators of sex differences in autoimmunity [53].

JOURNAL of MEDICINE and LIFE
Acquired MG may manifest for the first time during pregnancy [3]. Pregnancy has a variable effect on the disease; 41% of patients will present exacerbation, 29% will show remission, and no change will be present in the course of the disease of 30% [6]. Also, women often describe a worsening of the symptoms before their menstrual period, when progesterone levels are lowering [54].
Thymic hyperplasia affects female patients (ratio 9:1) during the fertile period of their life [55]. Estrogen receptors are expressed on thymic epithelial cells and thymocytes [52]. An increased expression of estrogen receptors was found on the thymocytes and T cells from peripheral blood mononuclear cells in MG patients [55]. Potentially, the B cells of MG patients show an aberrant reactivity to estrogens and an intensified B-cell response [55].
Androgens suppress both T-cell and B-cell immune responses, resulting in the suppression of disease expression serving as a protective factor [51]. The unimodal distribution among males might be explained by an age-related decline in testosterone [56]; however, this may also be coincidental.

Risk factors: genetic architecture, geo-epidemiology and the environment
MG is very rarely inherited, between 3.8% and 7.1% of MG patients reporting a family history of the disease [57]. Few cases of familial autoimmune MG have been reported: a family with parental consanguinity and five of 10 siblings affected by late-onset autoimmune MG [58], a Hungarian family where nine members from two generations developed MG [6], an Italian-American family with 5 children affected by early-onset MG linked to a variant in the ecto-NADH oxidase 1 gene (ENOX1) [6]. Monozygotic MG twin concordance is estimated at 35%, supporting a genetic influence in disease development [57].
Even though MG is rarely inherited, other autoimmune diseases, such as thyroid disorders and rheumatoid arthritis, are frequent among the MG patients and the myasthenic patient's relatives, suggesting that both genetics and the environment are predisposing factors [6]. Autoimmune thyroiditis occurs in 10% of MG patients [59].
Genetic susceptibility in patients with MG has been suggested in different human leukocyte antigens (HLA) associations. The correlation with HLA fluctuates widely in relation to sex, age at disease onset, and thymic histology of the patients, and it differs between Caucasian and Asian populations [60]. One important and consistent feature in MG is the association with different HLA antigens in different ethnic groups.
The HLA A1-B8-DR3-DQ2 haplotype, also known as AH8.1, is associated with early-onset MG in the Caucasian population [6], while late-onset MG is associated with a variety of other HLA markers [3]. Patients with MuSK-MG seem to be associated with HLA-DR and DQ5 alleles [6,42]. AH8.1 has also been associated with systemic lupus erythematosus (SLE) and celiac disease, suggesting a shared genetic background for autoimmune diseases [6]. The DQ9 and HLA-DRB1(*)09 haplotypes were associated with MG in Asian patients [6]. Although MG patients with diversified ethnic backgrounds carried different HLA haplotypes, several loci (HLA-B*08 and HLA-DQA1) presented virtually universally in certain MG subtypes. The similarities in HLA profiles conform to global human migration trajectories, during which the people with the same ethnic background carried common ancestral haplotypes [56].
Early-onset MG, which occurs more frequently among young women, is associated with HLA-A1, B8, -DR3 alleles, whereas late-onset MG occurs predominantly in men who demonstrate associations with HLA-A3, B7, -DR2 alleles [8]. Some alleles are suggested to be protective for men with MG [61].
In an individual with a susceptible genotype, exposure to environmental factors can act to initiate an autoimmune process [62]. The gene variants that make an individual susceptible or protect him from autoimmune disease development are probably a result of evolutionary adaptation to the environment [8]. Environmental factors, such as stressful life events, viral infections, and various drugs or toxins, have been postulated to precipitate the development of MG. Some patients report that the disease is precipitated by a viral or bacterial respiratory or other infection (4%), emotional stress (4%), a physical trauma (3%), hyperthyroidism (2%), or administration of thyroid hormones (1%), operative procedures especially thyroidectomy (1%), pregnancy or delivery (1%), allergic reactions (1%), or exposure to drugs such as quinidine, procainamide, penicillamine, aminoglycoside, or other antibiotics (1%) [39].
A correlation between autoimmune diseases and infections has been suggested, including hepatitis C, herpes simplex virus, Epstein-Barr virus, Cytomegalovirus, Human T-lymphotropic virus, West Nile Virus [46], and more recently SARS-CoV-2 [63]. SARS-CoV-2 may exacerbate MG [64] and also trigger it. Three cases of AChR antibody-positive MG associated with SARS-CoV-2 were described in August 2020 [63], with MG symptoms occurring within 5-7 days after the onset of fever in patients without previous neurological or autoimmune disorders. The "hygiene hypothesis" suggests that a better quality of life and a fewer incidence of infections like tuberculosis in the developed world explain the rise in incidence and prevalence of MG. However, data from countries with high rates of infectious diseases and similar MG trends do not uphold this theory [8].
A multitude of toxins from animal, vegetal, and bacterial sources are able to disturb the function of the NMJ. Insecticides used in agriculture act on nerve targets, such as acetylcholinesterase, pseudocholinesterase, and the AChR [65]. This leads to questioning whether recurrent exposure to pesticides could cause MG (by sensitizing the immune system to the cholinergic receptor) [65]. In this regard, several studies have been conducted in agricultural populations. One of them revealed an increased prevalence of MG in adult men from rural areas in Morocco, including workers exposed to pesticides suspected of triggering autoimmune diseases, suggesting a role of these substances [66]. Another study detected an increased prevalence of MG among rural male adults in Israel, associating it with pesticide exposure [65]. It cannot be ruled out that other factors may explain this association, but physicians might consider pesticide exposure when patients present with MG symptoms [65].

CONCLUSIONS
Epidemiological data regarding MG are important for postulating hypotheses regarding its etiology and facilitating the description of MG subtypes. Thus, adequate documentation through broad databases is essential.
The incidence and prevalence of MG reported around the globe have been rising steadily and consistently over the past decades. We expect the future MG patient to be older and have more comorbidities, which will lead to a more elaborate treatment.
Ethnic differences have been described, suggesting that genetic or environmental/lifestyle factors contribute to a specific phenotype, but also that MG may be considered as a complex of symptoms associated with different genetic susceptibility rather than a single clinical entity.
MG is still a disease of young women and older men. Gender is a deciding epidemiological risk factor for the development of autoimmune diseases, given that the immune system is a subject of sexual dimorphism. Many environmental factors are associated with MG, supporting the hypothesis that genetically sensitive individuals develop the autoimmune disease after exposure to environmental triggers.