Diagnostic delay in a case of herpes simplex encephalitis

Abstract

Herpes simplex encephalitis (HSE) is the most frequent cause of sporadic fatal encephalitis in the Western world. Definitive diagnosis by viral PCR of cerebrospinal fluid (CSF) and treatment with aciclovir have improved the prognosis significantly. Nevertheless, the condition is rare and presents with non-specific symptoms that can easily be mistaken for systemic infection or non-infective encephalopathy. We report a case of HSE which was not recognised by four separate doctors, leading to substantial delay in diagnosis and treatment. Our patient presented with fever, headaches, altered behaviour and generalised bradykinesia. This was initially diagnosed as otitis interna (labyrinthitis) and, subsequently, an ischaemic stroke. There was a delay of 10 days in the initiation of aciclovir from symptom onset. MRI and CSF PCR confirmed herpes simplex virus type-1 (HSV-1) infection. The patient improved on aciclovir, but is disabled with word-finding difficulties and cognitive slowing.

BACKGROUND

Herpes simplex encephalitis (HSE) is the most frequent cause of sporadic fatal encephalitis in the Western world, with an incidence of one case per million per year.¹ In 90% of cases, the cause is herpes simplex virus-1 (HSV-1), with HSV-2 causing the remaining cases, mostly in neonates and immunocompromised individuals.² Viral encephalitis classically presents with symptoms of fever, headache, altered behaviour, fluctuating consciousness, seizures and focal neurological deficit. In reality, many patients will not show all these features, and it can sometimes be difficult to distinguish the condition from metabolic encephalopathy. Important distinguishing features include fever, headache and focal neurological signs, all expected in encephalitis but absent in metabolic encephalopathy.¹

HSE usually displays a rapid onset with swift progression over days.³ However, there are no pathognomonic features in the history or the examination that may be used to distinguish HSE from other forms of infective encephalitis. The presence of oral cold sores is not a risk factor for HSE. Frontotemporal signs on neurological assessment, such as aphasia and personality change, should alert the doctor to the possibility of encephalitis.

Routine full blood count and biochemistry may be normal, as was evident in this case. Inflammatory markers may not be elevated, especially early in the disease process. MRI is the imaging modality of choice in cases suspected of viral encephalitis, and usually reveals hyperintense signal abnormalities on T2-weighted images in one or both inframedial temporal lobes, insular cortices, inferior frontal lobes, cingulate gyri and the thalamus.⁴ MRI is rarely normal in HSE.

Lumbar puncture is essential to diagnosis. A raised cerebrospinal fluid (CSF) lymphocyte count and protein level is expected. Advances in CSF PCR now permit the identification of HSV DNA with sensitivity and specificity >95%, rendering brain biopsy almost redundant.

Antiviral therapy in the form of aciclovir has revolutionised the treatment of HSE, reducing mortality from 70% in untreated cases to 19% and 28% at 6 and 18 months, respectively.⁵ Importantly, however, for patients treated with aciclovir, 25% are left with severe neurological impairment. Acting as a specific inhibitor of herpesvirus DNA polymerase, aciclovir has been shown to possess a remarkable safety profile.⁶ Nevertheless, for effective therapy, it is essential that the drug is initiated before the onset of haemorrhagic necrosis, an event correlating with significant deterioration in level of consciousness. The efficacy of aciclovir, as well as its safety profile have led to the recommendation that aciclovir be initiated as soon as infective encephalitis is suspected in a patient, with cessation of the drug should further investigation lead to an exclusion of viral encephalitis.

CASE PRESENTATION

A taxi-driver in his fifties presented to Accident & Emergency on a Friday accompanied by his wife, who complained that she had noticed a “general slowness” in her husband for 1 week. She explained that her husband had returned from work 7 days previously with a thumping headache, after which he fell asleep for almost 24 h. Since then, the patient had complained of fever, lethargy and anorexia. He also became unsteady on his feet and was “talking silly”, such that members of his family struggled to understand him.

He had seen his general practitioner (GP), who prescribed doxycycline for otitis interna (labyrinthitis). This failed to relieve his symptoms. The patient continued to experience intermittent headaches, which were not alleviated by simple analgesia. He became more unsteady on his feet, with bradykinesia and slurred speech, and general blunting of his affect. Three days after seeing his GP, he presented to hospital.

His past medical history is significant for type 2 diabetes mellitus, for which he was taking oral hypoglycaemics, as well as hypertension and high cholesterol. Also of note is a history of a congenital cranial meningocoele, which had been excised at 1 month of age and which had not given him any further cause for concern.

His father had an ischaemic stroke at the age of 68 years, and had also had a coronary artery bypass graft after a myocardial infarction. His mother has angina.

He works as a taxi-driver at an international airport, smoking 10–15 cigarettes per day and drinking alcohol rarely.

On examination, the patient was comfortable at rest with temperature 37.6°C. His Glasgow Coma Scale was 15/15, but he was markedly inexpressive. His pulse rate was 58 bpm and regular, with blood pressure 132/75 mm Hg. Heart sounds were normal to auscultation, with no added sounds and a normal jugular venous pressure. His respiratory rate was 14, oxygen saturation 94% on room air, and chest clear to auscultation. His abdomen was soft and non-tender.

Neurological examination revealed slow but otherwise normal gait, with slurred speech and generalised bradykinesia. There was no pronator drift. Examination of the cranial nerves found no abnormality of note, with visual fields full to confrontation. Tone, power and reflexes in all four limbs were normal and symmetric. Coordination, however, was deficient in the upper limbs, with poor finger–nose test bilaterally.

The patient scored 8 out of 10 on the Mental Test Score, losing marks for failure to recall an address and the years of the First World War.

Full blood count, biochemistry and glucose were all within normal limits. C-reactive protein, however, was mildly elevated at 28 mg/l. ECG revealed the patient to be in normal sinus rhythm.

CT head (fig 1) revealed evidence of mega cisterna magna, with a prominent ventricular system, in keeping with the patient’s history of congenital cranial meningocoele. Also of note was an area of hypoattenuation within the right frontal and temporal lobes which was reported as being consistent with an acute ischaemic cerebral infarction in the distribution of the right middle cerebral artery. A right frontoparietotemporal cerebrovascular accident (CVA) was diagnosed, and the patient was admitted to a combined acute and rehabilitation stroke unit.

Figure 1.

CT scan of the head. Transverse section at the level of the orbits. Note the region of hypodensity in the right frontal, parietal and temporal lobes, as well as the mega cisterna magna, a feature of this patient’s excised cranial meningocoele.

On the Monday (day 3 postadmission and 10 days after symptom onset), the patient was seen on an acute stroke ward round by a consultant stroke physician. The patient’s symptoms had remained static and an intermittent pyrexia had been documented. A diagnosis of encephalitis was considered. High-dose aciclovir (10 mg/kg three times a day intravenously) was started immediately, before an urgent lumbar puncture was performed. This atraumatic tap revealed the presence of clear CSF with an opening pressure of 120 mm H₂O (normal 60–250 mm H₂O). CSF laboratory analysis revealed a white cell count of 514×10^–6/l (99% lymphocytes) (normal <5×10^–6/l), erythrocyte count of 82×10^–6/l, protein at 2.90 g/l (normal <0.4 g/l) and a normal glucose level at 3.1 mmol/l (serum glucose 6.6 mmol/l) (normal two-thirds of serum glucose). CSF was sent for PCR.

An urgent MRI scan of the head, performed the same day, revealed diffuse hyperintensities in the white matter of the right frontal, parietal and temporal lobes, consistent with encephalitis (fig 2).

Figure 2.

T2-weighted MRI scans of the head. (A) Transverse section at the level of the optic chiasm demonstrating high signal in both white and grey matter of the right frontal, parietal and temporal lobes. (B) Coronal section, showing mass effect with compression of the right lateral ventricle.

After this, the differential diagnosis was revised to include HSE, as well varicella zoster. Given the patient’s occupation, more exotic infective encephalitides also needed to be considered, including West Nile fever, Japanese encephalitis, as well as Western and Eastern equine encephalitis. Tuberculosis was also considered. The patient was started on high-dose amoxicillin (3 g four times a day intravenously) as well as gentamicin (5 mg/kg once a day intravenously) to cover Listeria encephalitis.

OUTCOME AND FOLLOW-UP

Three days into antiviral therapy, the patient’s clinical picture improved, with less hesitant speech. HSV-1 DNA was detected in the CSF by PCR. Antibacterial chemotherapy was discontinued, with a plan to continue intravenous aciclovir for a minimum of 2 weeks pending clinical improvement, followed by 4 weeks of oral valciclovir.

Despite treatment with aciclovir and some clinical improvement the patient remains off work, suffering continued word-finding difficulties and cognitive slowing.

DISCUSSION

Benson and Swadron⁷ investigated the initiation of empirical aciclovir in the Emergency Department (ED) in patients ultimately diagnosed with encephalitis over a 10 year period. Their inclusion criteria encompassed patients presenting with fever, neuropsychiatric abnormality, and CSF pleocytosis with a negative Gram stain result in the ED. Of 24 patients meeting these criteria, only 7 (29%) received empirical aciclovir in the ED.

The non-specific and varied nature of presentation makes diagnosis of HSE challenging in both adults^8–11 and children.¹² In 1985, before the routine use of PCR for diagnosis and aciclovir for treatment of HSE, Tashjian and Lane presented two cases of patients with the condition whose symptoms were incorrectly attributed to other diagnoses.⁸ In the first patient, postoperative fever and grand mal seizures led to a diagnosis of nosocomial infection, whereas in the second, upper limb weakness and seizures were attributed to alcoholic hepatic encephalopathy. Both patients suffered grave outcomes. Furthermore, Townend and colleagues⁹ highlight a case in which the diagnosis of HSE was missed in a patient presenting in a more textbook manner. Symptoms of fever, confusion and speech disturbance, together with positive CT findings were initially attributed to pneumonia in a patient with stroke as HSE was not included in the differential diagnosis.

Two groups of patients have been shown to have poor outcome after HSE—those presenting with severe symptoms and signs, and those in whom treatment with aciclovir is delayed by >2 days after admission.¹³ The patient described in this case report was seen by four doctors (GP, A&E doctor, junior hospital doctor and consultant on-call), all of whom failed to suspect encephalitis. He was eventually started on aciclovir on day 3 as an inpatient, and 10 days after the onset of symptoms. A high index of suspicion for a diagnosis other than ischaemic stroke should be held when a significant volume of hypoattenuation on cranial CT is accompanied by paucity of lateralising neurological deficit on clinical examination.

LEARNING POINTS

• Herpes simplex encephalitis is associated with 70% mortality if untreated.

• Delay in treatment leads to higher morbidity and mortality.

• The diagnosis must always be considered in a patient presenting with abnormal behaviour.

• Robust correlation of imaging studies with the clinical picture is paramount.

• Once suspected, high-dose aciclovir must be started immediately, before lumbar puncture, and only stopped once a definitive alternative diagnosis has been made.

• MRI and CSF PCR for viral DNA are essential for definitive diagnosis.