Hepatitis E virus (HEV) is endemic in many developed countries, with seroprevalence rates ranging from 4.7% in Scotland to 31.1% in France. Cases are caused by genotypes 3 and 4, which are porcine zoonoses (Figure 1). Infection is associated with the consumption of undercooked pork meat and is often asymptomatic.1 In England there are an estimated 100,000 new infections each year, but <1% of cases are laboratory-confirmed.2 A significant cause of this disparity is the currently incompletely understood clinical phenotype of HEV infection.
HEV infection has been associated with a variety of extra-hepatic manifestations, the most common of which are neurological.3,4 To date, ~100 cases of hepatitis E-associated neurological injury have been described in case reports/series.5 HEV infection has been linked to a number of neurological conditions including neuralgic amyotrophy (NA), Guillain-Barré syndrome (GBS), and meningoencephalitis. Recent studies have found that 5–11% of GBS patients6 and 10% of NA patients7 had evidence of HEV infection at the onset of their neurological illness. Importantly, in these cases neurological signs and symptoms dominated the clinical picture; patients were mostly anicteric and had normal or only mildly elevated liver enzymes. The diagnosis of HEV is easily, and probably commonly, overlooked.
To further elucidate the range and frequency of HEV-associated neurological injury we conducted a multicentre study across four centres in the UK, France, and the Netherlands. Four hundred and sixty-four consecutive patients who presented with acute non-traumatic neurological illnesses were prospectively tested for HEV by serology and polymerase chain reaction (PCR). Of these, 11 patients (2.4%) had evidence of current or recent HEV infection, 7 (1.5%) of whom were HEV PCR positive. Neurological presentations were wide ranging and included NA (n=3, all PCR positive), cerebrovascular event (n=4), seizure (n=2), encephalitis (n=1), and an acute combined facial and vestibular neuropathy (n=1). In all cases, the hepatitis was mild and no patients were jaundiced. All strains were classified as genotype 3. The cases of NA were all middle-aged males with bilateral involvement of the brachial plexus.
The most important question regarding the above observations is: ‘Is the relationship between HEV infection and neurological injury just a chance association or is the relationship causal?’ In NA, our view is that this relationship is causal. Our reasoning for this assertion is based on very recent findings from another multicentre European study of 118 patients with NA, ~50% of whom were associated with HEV infection.8 This study showed that bilateral involvement of the brachial plexus is the clinical hallmark of HEV-associated NA. The cases of NA we observed in the current study all had bilateral involvement of the brachial plexus. This observation, together with a temporal relationship between HEV infection and NA in differing geographical locations in Europe, together with laboratory data which show that HEV is potentially neurotropic,9 is strong evidence supporting a causal relationship between HEV infection and NA.
Our finding that 1.5% of subjects with acute non-traumatic neurological injury have HEV viraemia at presentation is striking but needs to be contextualised. The prevalence of HEV genotype 3 viraemia among asymptomatic blood donors ranges from 0.035% in the UK (2013–4),2 0.045% in France (2012–3), and 0.17% in the Netherlands (2013–4).10 Although the HEV viraemia rate in neurological cases we observed is an order of magnitude higher than we would have expected from these blood donor studies, this by itself does not prove causality. The reason for this is that the incidence and prevalence of HEV infection in developed countries varies both geographically and over time. Careful case-control studies are now needed to further determine the role of HEV in patients with neurological injury, other than NA.