Fatal neonatal parechovirus encephalitis

Abstract

Two infants developed encephalitis in the late neonatal period due to human parechovirus type 3 (HPeV-3). This finally resulted in intractable seizures leading to death. Both presented with classical signs and symptoms. HPeV-3 was detected in nasopharyngeal and rectal swabs, cerebrospinal fluid, serum and postmortem samples (for one infant). Ultrasound and magnetic resonance imaging findings showed severe white matter injury in one infant, and limited hippocampal injury in the other. Parents consented to autopsy in the latter, showing diffuse gliotic oedema of cerebral white matter, bilateral bronchopneumonia, and inflammation in spleen. We emphasise that polymerase chain reaction (PCR) testing for parechovirus in neonates with severe sepsis-like syndrome and meningoencephalitis is mandatory.

Background

Human parechovirus, a Picornavirus family member, has 14 known serotypes. They are as ubiquitous as enterovirus and the first contact is usually within the first months of life.¹ Human parechovirus type 1 and 3 are the most frequently detected serotypes. Symptoms are fever, rash, gastroenteritis and upper respiratory tract infection. HPeV-3 in particular is a known cause of sepsis-like syndromes and encephalitis in young children, though death has to our knowledge not yet been reported. Earlier reports showed diffuse white matter injury on magnetic resonance imaging (MRI) and ultrasound.² We present two fatal cases of HPeV-3 encephalitis, only one with extensive white matter injury.

Case presentation

Case 1

A girl with a birth weight of 2800 g, born at 38 weeks, presented with lethargy, poor feeding, vomiting and a temperature of 38.4°C on day 21 of life. She developed generalised seizures. She was irritable, had a bulging fontanel, a generalised erythematous macular rash, vertical nystagmus and hypertonia, especially of the lower extremities. Laboratory results showed a maximum C reactive protein (CRP) value of 41mg/l. Cerebrospinal fluid (CSF) could not be tested on levels of glucose and protein and cell count, due to lack of material. Serial ultrasound of the brain demonstrated hyperechoic change in frontal white matter, gradually increasing and leading to cavitation 1 week after admission. Hyperechoic change in the thalamus was observed during this evolution. MRI on day 24 confirmed white matter injury by showing hypersignal on diffusion weighted imaging (DWI), especially in the frontal and parietal lobes (fig 1). A follow-up MRI on day 33 confirmed cavitation in the areas of white matter affected earlier. An electroencephalogram (EEG) documented multifocal epileptic activity. Management consisted of endotracheal intubation, broad spectrum antibiotics, aciclovir, ganciclovir and different anticonvulsant drugs (phenobarbitone, midazolam, lidocaine, pyridoxine, pentobarbitone). Because of therapy resistant status epilepticus due to severe irreversible brain damage, ventilation was withdrawn on day 37 of life.

Figure 1.

Diffusion weighted magnetic resonance imaging on the third day after presentation in patient 1. Inset: coronal reconstructed diffusion trace image, red colour highlighting areas with restricted diffusion. Structures affected are frontal and temporal white matter, cerebellar cortex, posterior limb of the internal capsule and genu corporis callosi, and anterior and mediodorsal thalamic subnuclei with the anterior thalamic radiation in the anterior limb of the internal capsule.

Case 2

A boy of 2210 g at birth, firstborn of a twin pregnancy at 34 weeks, presented with coughing, poor feeding and progressive respiratory distress on the 35th day after birth. He was drowsy, with a temperature of 39°C. He had a generalised erythematous macular rash, seizures, hypotonia and fasciculations of the tongue. His twin brother had no signs of infection and was not tested for HPeV-3. Laboratory results showed a maximum CRP value of 26 mg/l. CSF showed normal cell count and glucose. Protein concentration was slightly increased (0.4 3g/l). An initial ultrasound excluded congenital anomalies, focal lesions and white matter change. MRI showed focal hypersignal on the diffusion map in hippocampus and fornix (fig 2). An ultrasound scan 5 days later demonstrated discrete ventriculomegaly with mildly dilated extracerebral fluid spaces. On EEG a high index of multifocal epileptic activity was observed, corresponding with the fasciculations. Management consisted of endotracheal intubation. He was treated intravenously with a broad spectrum antibiotic. Subsequently, various anticonvulsant drugs were administered (midazolam, phenobarbitone, lidocaine, pyridoxine, pentobarbitone). None were successful. Due to therapy resistant status epilepticus, treatment was withdrawn. He died on day 53 of life. Autopsy showed a brain with a weight of 402 g (normal 506 g), with signs of severe meningitis and hypoxia. There was diffuse gliosis and oedema of white matter in the cerebrum and cerebellum. HPeV-3 was isolated in postmortem samples of brain and lung tissue. Autopsy also showed pus in the trachea, bilateral bronchopneumonia and inflammation in spleen.

Figure 2.

Coronal ultrasound scans and diffusion weighted magnetic resonance imaging in patient 2 on the eighth day after presentation. Normal echoic aspect of white matter and thalamus, mild ventricular dilatation. Hypersignal on diffusion weighted imaging symmetrical in hippocampus and fornix (arrow), unilateral in right caudate head and in a limited area of right frontal white matter.

Investigations

In both patients serum, faeces, urine, nasopharyngeal aspirate and cerebrospinal fluid were collected. In all but the urine specimen HPeV-3 was detected using PCR and VP1 sequencing. The samples were negative for other viruses and bacteria.

Discussion

These two cases illustrate the range of effects of HPeV-3 on the neonatal brain. Death due to HPeV-1 has been described in two previous case reports.^3,4 To our knowledge death due to HPeV-3 has not been reported.

HPeV-3 infections are common. Up to 80% of the adult population has serological HPeV-3 antibodies.¹ Most infections take place in early childhood. Recent studies showed HPeV-3 was more involved in central nervous system (CNS) pathology than other parechoviruses, especially in neonates.⁵ Wolters et al showed that HPeV specific polymerase chain reaction (PCR) testing led to a 31% increase in detection of a viral cause of neonatal sepsis or CNS symptoms in children under the age of 5 years, making HPeV-3 the second cause of viral sepsis and meningitis in young children.⁶ With this in mind, and the finding that human parechovirus can cause severely life threatening disease, it is obvious that parechovirus infection should be queried when facing meningitis or encephalitis in young infants. Although in our cases the diagnosis was made shortly after admission, this could not prevent fatal outcome. Things may change when treatment options become available.

Two questions come to mind when comparing ultrasound and MRIs of these patients. First, our second patient had only minor lesions on his ultrasound and MRI scan compared to our first patient. When comparing the MRI findings with earlier published results, the lesions of patient 2 seem to have a different pattern.² Verboon-Maciolek et al described 10 cases of neonatal encephalitis due to HPeV-3. In nine cases an ultrasound was performed showing severe periventricular echogenicity in all. In eight cases diffusion weighted images were available; all eight showed mild or high signal intensity of the periventricular white matter and corpus callosum. These findings correlate with ultrasound and MRI findings in patient 1 but not patient 2. Patient 2 only had bilateral lesions in hippocampus and fornix, the right head of nucleus caudatus, and a limited region of right frontal white matter. Repeated ultrasound in patient 2 showed no abnormal echogenicity. Whether these apparent differences have any pathogenetic meaning remains unclear. Perhaps it is due to different methods of brain invasion, but other factors could also have contributed. Perhaps the findings in patient 2 are consequent to seizures or periods of hypoxia. Second, though minor lesions were seen on in vivo images of patient 2, the postmortem examination did, however, show diffuse gliosis and oedema of the white matter of the cerebrum and cerebellum. Perhaps a time lag between MRI and death offers an explanation, although the ultrasound did not change much during admission. His autopsy report stated that there were signs of meningitis and encephalitis, there was bilateral bronchopneumonia, and the spleen also showed signs of infection. Samples of brain and lung tissue tested positive for HPeV-3.

Both patients showed similar clinical signs. Apart from the status epilepticus, these signs are seen in most neonatal HPeV-3 infections. Laboratory findings in our patients did not differ from findings published earlier.⁶ It remains unclear why HPeV-3 was fatal in these particular cases, while in the large majority of cases it is a self limiting disease.

Based on this we can only state that the correlation between clinical signs, laboratory results, radiological findings and histological findings needs further study.

Learning points

• Human parechovirus type 3 (HPeV-3) can cause severe life threatening disease and even death.

• HPeV-3 should be regularly tested for in infants with signs of sepsis-like syndrome and meningoencephalitis.

• Radiological findings do not always correlate with clinical presentation and autopsy findings.