Non-traumatic neurological injury and hepatitis E infection

KEYWORDS:

• Autochthonous
• Guillain–Barré syndrome (GBS)
• hepatitis E virus (HEV)
• neuralgic amyotrophy (NA)
• neurological injury

1. Introduction

The association between viruses and neurological diseases has long been documented in the medical literature. However, our understanding of viral-induced neurological injury remains limited. This is reflected by the sheer number of neurotropic viruses, the wide spectrum of clinical presentation and the complex interaction between pathogen and host immunological response [1]. A recent prospective multicenter study from France, the Netherlands, and UK showed that in 464 consecutive patients presenting to hospital with nontraumatic neurological injury, 11 (2.4%) had evidence of current or recent hepatitis E virus (HEV) infection [2]. In this editorial, we present current evidence supporting the link between HEV and neurological injury, and question its taxology as solely a hepatotropic virus.

2. Hepatitis E

Since its discovery almost 30 years ago, hepatitis E has been previously somewhat neglected by the scientific community. Traditionally, the single-stranded, positive sense RNA virus was considered to be restricted to a number of developing countries in Asia and Africa, where it is a waterborne infection caused by HEV genotypes (gt) 1 and 2, which are obligate human pathogens. Until relatively recently, HEV infection in developed countries was considered a ‘back-packers’ disease and confined to travelers returning from endemic regions [3,4]. However, it is now clear that locally acquired HEV is also endemic in developed countries [3,4].

In developed countries, hepatitis E is caused by gt3 and 4, which are primarily zoonotic diseases with pigs as the most important host. Although HEV is nonpathogenic in pigs [5], in humans it causes acute hepatitis predominantly in older males, and chronic hepatitis in the immunosuppressed including transplant recipients [3,4]. One important route of infection from pigs to humans is via the consumption of HEV-infected pork-meat products. The amount of circulating HEV in humans in Europe varies between and within countries and over time. HEV is very common in some western European countries including the Netherlands, Germany, U.K., and particularly France where (for unknown reasons) it is hyperendemic in the southeast and southwest of the country. In recent years, the number of laboratory-confirmed cases has increased rapidly in these locations. HEV has also found its way into the human blood supply, with high rates of viremic donors in France (1:2218), UK (1:2848), and the Netherlands (1:600) [6–8]. This has recently prompted the introduction of universal donor-screening in the latter two countries. These data confirm that huge amounts of HEV are circulating in these populations, most of whom have asymptomatic infection. Only a small minority (<5%) develops symptomatic hepatitis on exposure, but an unknown number are symptomatic, but currently not recognized. This includes patients who present with primarily a neurological illness.

3. Hepatitis E and neurological injury: current evidence

HEV-associated neurological injury has been documented in acute and chronic infection and in developing countries (gt1 endemic areas) and developed countries, but is best documented in Europe in the context of zoonotic HEV gt3 infection. HEV-associated neurological injury affects both the central and peripheral nervous systems [9,10] and has been reported in Guillain–Barré syndrome (GBS), neuralgic amyotrophy (NA), encephalitis/myelitis, mononeuritis multiplex, myositis, peripheral neuropathy, vestibular neuritis, and Bell palsy [9]. In the neurological community it has been known for a good number of years that some cases of GBS [11] and NA [12] have unexplained liver enzyme abnormalities at presentation, which was not possible to correlate with any known cause of hepatitis. However, at that time, these patients were not tested for HEV. We now know that the strongest association between HEV infection and neurological damage is with inflammatory neuropathies such as GBS, and NA and also meningoencephalitis.

Fifty-three cases of HEV-associated GBS have been reported (mostly from Europe), including a range of GBS variants. Anti-ganglioside antibodies have been documented in 8, and HEV RNA found in serum of 17 of these patients. Case-control studies show that serological and/or molecular evidence of HEV infection is found in 5%–11% of cases of GBS, significantly higher than in controls [13–15]. Over 70 cases of HEV-associated NA have been documented almost exclusively from Europe with HEV gt3. A recent multinational retrospective study of 118 patients with NA (approximately 50% of which were HEV-associated) found that patients with HEV-associated NA (HEV-NA) have a distinct clinical phenotype, with bilateral involvement of and more extensive damage to structures within and outside the brachial plexus, including the phrenic nerve(s) [16]. HEV-NA patients are generally viraemic when tested within 2 weeks of symptom onset and HEV RNA has been found in CSF one case so far. Thirteen cases of HEV-associated meningoencephalitis have been reported, both in acute and chronic infection. HEV RNA is commonly found both in serum and CSF in the immunosuppressed who sometimes have an ataxic component to their symptomatology and appear to have a worse prognosis [10].

4. Pathogenic mechanisms and causality

The pathogenesis of HEV-associated neurological injury is uncertain. In GBS and NA, it could be due to immune-mediated mechanisms, congruent with existing notions in these diseases. However, HEV is not infrequently found in the serum and occasionally in CSF suggesting that HEV might be directly neurotropic, particularly in patients with meningoencephalitis and possibly in NA. This notion is given support by in-vitro studies which show that HEV can grow on neurological cell lines and can cross the blood brain barrier in animals [17,18]. A causal relationship for HEV is strongest in GBS, NA and meningoencephalitis. The notion of causality is supported by a number of observations, including the homogeneity of cases over time and geographical location, case-control data (GBS), quasispecies compartmentalization and intrathecal antibody synthesis, and resolution of symptoms with viral clearance [10]. In NA causality is additionally supported by the finding that 3/5 patients tested in the recent prospective study had HEV infection, and all had bilateral involvement, which we now know to be the dominant HEV-NA phenotype [2]. This prospective study also found a small minority of patients with stroke and seizures had evidence of HEV infection at presentation, but the role of HEV is such conditions remains to be established. Given the large amount of circulating HEV in the human population it could just be a chance association. To tease this out, prospective studies with appropriate controls are warranted. The optimal treatment for HEV-associated neurological injury is not yet defined. There have been a few case reports/series of the use of ribavirin, prednisolone and intravenous immunoglobulin, with mixed results, and randomized controlled studies are required [10].

5. The future

In our view, all patients with NA, GBS and encephalitis should be tested for HEV, and patients with other neuroinflammatory conditions should be considered for testing if they have unexplained abnormalities in liver function tests (LFTs). The full range and frequency of HEV-associated neurological injury is unknown and case-control studies are required. A recent epidemiological study (Neurocave) on the role of HEV in neurological conditions in southern Switzerland has just finished recruitment, and will hopefully shed further light on these issues.

Currently, we do not know the proportion of patients with HEV infection who present with primarily a neurological illness, compared to those who present with hepatitis. What we do know, however, is that patients with HEV-associated neurological injury rarely present with jaundice and the LFTs are not uncommonly normal [9]. This raises the issue of whether this intriguing virus has been misnamed. Perhaps it should be considered as ‘neuro-hepatitis E’?

Declaration of interest

H Dalton has had travel and accommodation costs and consultancy fees from GlaxoSmithKline, Wantai, and Roche; travel, accommodation and lecture fees from Merck, Gilead and GFE Blut GmBh; travel and accommodation fees from the Gates Foundation and Médecins Sans Frontières and a grant from BMA. V E Jeroen has had consultancy and/or lecture fees from Merck, Genzyme, Biogen, Novartis and Teva. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed. Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Acknowledgments

The authors thank Dr. Claudio Gobbi (Neurocenter of Southern Switzerland) for helpful comments and suggestions.

Additional information

Funding

The manuscript was not funded.