Is West Nile Virus infection associated with myasthenia gravis?

Introduction

West Nile Virus (WNV) is a single-stranded RNA virus of the Flaviviridae family which has been known since 1937. ¹ It causes a worldwide endemic and sometimes epidemic mosquito-borne infection. In 80% of cases, the disease is asymptomatic, in 20% it is mild and known as West Nile fever and in <1% it is a neuroinvasive disease with severe and sometimes fatal complications. ² Severe forms are characterised by acute neurological manifestations, usually encephalitis or meningoencephalitis which end fatally or cause polio-like acute flaccid paralysis or movement disorders. Because of difficulties in detecting cases with mild sequelae such as slight neurological and functional disorders, information on long-term outcome is limited and not yet clear. ³

Myasthenia gravis (MG) is an autoimmune disease characterised by disorders in signal transmission due to the binding of autoantibodies to post-synaptic proteins which are responsible for signalisation at the neuromuscular junction. It is clinically seen as fluctuating skeletal muscle weakness. Many patients have generalised disease while 10% have only ocular involvement. Anti-acetylcholinesterase antibodies are positive in 80–85% of generalised MG patients.^4,5 The aetiology is thought to be multifactorial, giving an impression of interaction between genetic and environmental factors. Viral infections are particularly blamed as autoimmunity triggering agents. ⁴ Many viruses, such as Epstein–Barr virus, cytomegalovirus and human foamy virus, are reported probably to have an association with MG. ⁵

Although some neurological manifestations have been described in association with WNV infection, MG was not reported with this disease until 2005. ⁶ In 2012, this suggestion was mentioned again. ⁷ In January 2014, Leis et al. reported six patients who developed MG in 3–7 months after WNV infection. ⁸ None of the patients showed any sign or symptom of MG beforehand and all were diagnosed clinically and confirmed by serology during their follow-up period. However, this knowledge is insufficient to comment about this interesting association. Consequently, we decided to investigate past WNV infections among the patients in our hospital with MG.

Patients and Methods

Our hospital is one of the largest tertiary care hospitals in Turkey, accepting patients from all over the country, especially Central Anatolia. The Neurology Department has been following nearly 100 MG patients for many years, and those patients with confirmed MG were selected for inclusion.

Our protocol conformed with the ethical guidelines of our institution. Informed consent was obtained from each participant. All were interviewed about the time of their illness and past history of encephalitis/meningitis.

Patients were excluded from the study if they had any viral infective symptoms in the 3 weeks prior to sampling, to avoid the possibility of false positive ELISA results.^9–11 Patients under intravenous immunoglobulin (IVIG) treatment were also excluded, as this is known to alter immunoglobulin levels.^12,13

Hepatitis C infection was ascertained from the records or tested for if this had not been done beforehand.

Other flavivirus exposure (Dengue, Japanese encephalitis, etc.) are not relevant as they are not endemic in Turkey.

Age- and sex-matched people among other follow-up patients, who would not have been excluded by the above criteria and who had no neurologic deficit or disorders, were chosen as a control group.

Samples were centrifuged at 3000 rpm for 10 min and separated sera were saved in −80℃ for 3 months before the assay study. The ELISA test (DRG® Flavivirus (West Nile) IgG ELISA (EIA-4400), DRG International, Inc., USA) was used to detect WNV IgG in serum samples. The ELISA test kit had been tested by using CDC West Nile ELISA as a comparator and it had 100% positive agreement and 96% negative agreement on 476 samples. A plaque reduction neutralisation test was planned if there were positive results, but no such positive results were obtained.

The SPSS 20 packet programme was used for statistical analysis. A P value >0.05 was suggested for statistical significance.

Results

A total of 96 such serologically confirmed MG patients were included in our study (Figure 1). Nine patients were unreachable and 28 refused to participate and/or provide serum samples. OPEN IN VIEWER

Five of the remaining 59 participants were excluded because they had fever or symptoms of viral infection at the time of blood sampling. Four patients were excluded because they were unavailable to discontinue IVIG therapy in the sampling period.

Finally, 50 patients were available for evaluation in the study (Figure 1).

Blood samples were taken from all 50 patients and 38 controls.

Of the 50 available patients, 22 were men (Table 1). MG had been present for a mean of 8.3 years and 36/50 had it for up to 10 years; seven had lived for 10–20 years with MG and seven had lived with it for >20 years (Table 2). None declared a central nervous system infection or vaccination for flavivirus in the past. No samples had positive ELISA results.

OPEN IN VIEWER

Table 1.

Comparison of demographical data and test results.

	Myasthenia gravis (n = 50)	Control (n = 38)	P value
Age (years)	50.1 ± 16.6	44.9 ± 13.5	0.1
Sex (Male/Female)	22/28	22/16	0.3
Disease duration (years)	8.3 ± 6.7	–
WNV IgG positivity	0	0	0

OPEN IN VIEWER

Table 2.

Distribution of MG patients according to disease time.

Disease duration (years)	n (%)
0–5	24
5–10	12
10–20	7
>20	7

Discussion

Viruses are commonly held to be triggers for autoimmune disease. Not much is known about how WNV may trigger autoimmune mechanisms although there are some studies.

WNV infection is known to lead many neurological complications of a wide variety.^6,14 Neuromuscular manifestations were among these but no case of MG was reported until Leis and Stokic confirmed three newly diagnosed MG patients who were recently diagnosed as having WNV infection and two cases of association between WNV infection and MG.^6,7 Leis et al. subsequently published a similar article on this issue ⁸ in which they further reported six WNV fever patients who developed MG following a 6-month period. ⁸

Thus we sought to confirm such an association by serologic tests in our MG patients. We examined patients of different age groups and variable disease duration times. Our control group also consisted of people of different ages.

We used ELISA for the detection of IgG as evidence of past WNV infection as it is the most used and most sensitive for this purpose. We planned to perform plaque reduction neutralisation test in positive samples for confirmation but neither the myasthenic patients nor the control group had any positive results.

A supposition would be that lack of WNV seropositivity among MG patients may be just because of the rarity of WNV seropositivity in the general Turkish population, which is estimated to be up to 1.6% by ELISA methods.^15,16 Leis et al. report the prevalence of MG among all neuroinvasive WNV cases from 2002 to 2011 in Mississippi as 1.2% (5:412) which is 100 times higher than the estimated prevalence (0.014–0.02%) of MG in the general population. ⁸

It may be that immunisation to WNV could abate with time and thus IgG might disappear in serum samples some time after viral exposure. There is not much knowledge on this point, but it is known that WNV induces a significant antibody response that remains present even 5 years after infection. Moreover, IgG levels at this time show fairly minimal decrease compared with paired levels at 6 months post infection. ¹⁷ There is no information about the definite duration of or levels of WNV IgG in a time course greater than 5 years because this has not yet been studied. Nonetheless, after much more than 5 years, a trend can be predicted as there is a gentle decrease trend between the 6th month and 5th year. There are also studies showing long duration times of antibody responses against other members of Flaviviridae.^18–20 For example, the majority of military personnel exposed to Dengue Virus during World War II, as well as volunteers exposed experimentally to the virus, had detectable antibody levels 35–60 years after infection.^18,19 Similarly, up to 97% of military personnel vaccinated with yellow fever vaccine during World War II were shown to have neutralizing antibody titres up to 35 years after vaccination. ²⁰

Nearly half of our patients had borne MG for up to 5 years and a two-thirds majority up to 10 years. Only 14% had the disease longer than 20 years (Table 2). Thus, evidence of WNV infection should have been detectable at least in the former group, if it had been present.

Our study group of 50 patients may be deemed representative as the prevalence is only 20:100,000 (0.014–0.02%) in America, ²¹ 10:100,000 in Croatia ²² and 35:100,000 in Sardinia. ²³

Thus we are confident of the reliability of our results concerning exposure to WNV in our MG patients. Our results show conclusively no single case of past WNV infection in our MG patient population. We do not consider that those excluded would have had a potential to change our results. Thus, our findings seem to contradict those of Leis et al. They described MG in six patients who had neuroinvasive WNV infection. None of them had a benign form of WNV infection. ⁸ None of our MG participants had in corroboration described or documented neuroinvasive infection in their past. This could be suggested as an explanation why none of our MG patients showed evidence of WNV exposure. It may be that only neuroinvasive disease is responsible for a relationship to MG. However, such a reasoning would necessitate accepting a pathological mechanism other than a cross-reaction between WNV proteins and acetylcholine receptor subunits.

Studies focusing on immunological mechanisms especially in neuroinvasive WNV infection may subsequently cast light on this question.