Recurrent brainstem encephalitis associated with herpes simplex virus type 1 DNA in cerebrospinal fluid

Herpes simplex viruses (HSV) are the most frequent cause of acute sporadic viral encephalitis in the United States. ^[1] Diagnosis of encephalitis, and other neurologic syndromes due to HSV, has been hampered by the difficulty in isolating virus from CSF. Early and accurate diagnosis of HSV encephalitis has been greatly facilitated by using polymerase chain reaction (PCR) to identify HSV DNA in CSF from infected individuals. ^[2-5] This technique has also implicated HSV as a major etiologic agent in neurologic syndromes previously thought to be idiopathic, such as benign recurrent lymphocytic (Mollaret's) meningitis ^[6-8] and recurrent ascending myelitis. ^[9]

HSV may also be an important cause of viral brainstem encephalitis (BSE). ^[10-15] Reports of this syndrome, with rare exceptions, ^[11,15] describe fatal cases. Diagnosis has typically depended on detection in postmortem brainstem tissue of Cowdry type A intranuclear inclusion bodies in neurons and glia, ^[14,16] herpesvirus-like particles on electron microscopy, ^[14] and HSV antigen by immunocyto-chemistry. ^[10,16,17] In isolated cases, HSV has been cultured from CSF ^[10,15,18] or brainstem tissue. ^[18]

In situ hybridization (ISH) studies on post-mortem brainstem tissue of patients who died from microglial nodular BSE demonstrated the presence of HSV nucleic acids in approximately 50% of cases. ^[19] However, ISH studies, which require brainstem tissue, cannot be used for antemortem diagnosis of BSE. PCR amplification of DNA in CSF can be performed rapidly and safely ante mortem. We now describe the use of PCR to detect HSV DNA in CSF from a patient with recurrent BSE and suggest that PCR amplification may be of value in the antemortem diagnosis of BSE.

Case report.

A 47-year-old man with partial complex seizures since age 17 developed retro-orbital pain, headache, and posterior neck stiffness in February 1990. He complained of diplopia, tingling paresthesias in the left hand and leg, and associated left-sided clumsiness. An ophthalmologist discovered that he had impaired upward gaze. He had a single herpetic lesion on the lip. On neurologic examination, his mental status was normal, although he complained of feeling confused and irritable. Cranial nerve examination showed equal-sized, reactive pupils, full visual fields, normal acuity, and benign fundi. He had full and conjugate lateral gaze and normal downward gaze but markedly limited upward gaze. Nystagmus was noted on lateral gaze. He had decreased sensation to pinprick over the left cheek. Corneal reflexes were intact. Facial strength was normal. The lower cranial nerves were intact. He had normal strength and muscle tone, but a mild tremor in the left arm with maintained posture. Deep tendon reflexes were normal. Plantar responses were flexor. He had decreased sensation to pinprick over the left arm and diminished position sense. There was no incoordination, ataxia, or dysmetria. Gait was normal and Romberg test was negative. MRI of the brain was unremarkable. CSF analysis showed an opening pressure of 220 mm H₂ O, 13 cells/mm³ (100% mononuclear), protein 76 mg/dl, and glucose 75 mg/dl Table 1. Viral, bacterial, and fungal cultures, antigen tests, and VDRL of CSF were negative. Routine hematologic studies, including ESR, and blood chemistries were normal. He received prednisone, initially 80 mg/d and subsequently tapered. His symptoms resolved, but recurred approximately 7 weeks later. MRI of the brain was again normal. Repeat CSF examination showed an opening pressure of 195 mm H₂ O, 3 mononuclear cells/mm³, protein 71 mg/dl, glucose 64 mg/dl. He was restarted on prednisone, and symptoms again resolved. Over the next 18 months he had multiple recurrences of symptoms, often in association with the appearance of labial or perioral herpes simplex. On three occasions during this period (twice in July 1991 and once in September 1991) he received empiric therapy with oral acyclovir (1,000 mg/d times 5 to 10 days). Results of subsequent CSF examinations are shown in the Table 1.

HSV DNA was first detected in the CSF on 10/8/91 using previously described methods ^[6,7] Figure 1. Restriction typing of the amplified DNA, based on the presence of Alu I and Bal I restriction enzyme polymorphisms between HSV-1 and HSV-2, identified the amplified DNA as type 1. ^[6,7] Comparison of the HSV DNA primer and probe sequences used with other known DNA viruses indicates complete specificity for HSV-1 and HSV-2 DNA detection. Control CSF specimens known to be PCR positive for cytomegalovirus (CMV) and varicella-zoster virus (VZV) were negative in the HSV PCR assay (data not shown). CMV and VZV DNA were not detected in CSF by PCR when using CMV- and VZV-specific primers. HSV-1-specific CSF antibody ^[6,7] was absent on 10/8/91, weakly positive on 10/18/91, and strongly positive on 2/15/92 and 2/21/92 Table 1. HSV antibody was found in serum from both 10/8/91 and 10/18/91 samples. Antibodies against VZV and CMV were not detectable in CSF (less than 1:2, latex agglutination assay), although serum VZV antibodies were present (1:64, latex agglutination assay). These results suggest that there was specific intrathecal synthesis of HSV-1 antibody and are indicative of acute CNS HSV infection. As a result, on 10/8/91 he received another 10-day course of oral acyclovir (1,000 mg/d), with improvement in his condition. HSV DNA could not be detected in CSF obtained immediately following acyclovir therapy Figure 1.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1. Ethidium bromide-stained 3% agarose gel (NuSieve 3:1 agarose, FMC Corp., Rockland, ME) of DNA amplification products. CSF specimens are indicated by date of lumbar puncture. Water was used as a negative control; the positive control was 20 molecules of the plasmid pGX146. ^[7,31] Negative controls included CSF known to be PCR-positive for CMV and VZV (not shown). The specificity of the 224-base-pair bands was confirmed by blotting of the gel and hybridization with the HSVPolAInt probe (data not shown). ^[6] The primer and probes employed are completely specific for HSV-1 and HSV-2 DNA and do not cross-react with other known DNA viruses.

Three weeks later his symptoms recurred. He was restarted on oral acyclovir (600 mg/d) with remission of symptoms for the next 2 months. In February 1992, his symptoms recurred despite maintenance acyclovir therapy. He had decreased upward gaze and a very mild left hemiparesis involving the face, hand, arm, and leg. His left-hand rapid alternating movements were slower and clumsier than on the right. Deep tendon reflexes were not increased and the plantar responses remained flexor. Position sense was decreased in the left hand and pinprick sensation was diminished in the left arm. He was mildly dysmetric on left finger-to-nose maneuver. He had moderate nuchal rigidity and associated pain on neck flexion. CSF at that time showed 100 cells/mm³ (68% polymorphonuclear neutrophils, 32% lymphocytes), protein 61 mg/dl, and glucose 70 mg/dl. PCR amplification was positive for HSV-1 DNA on 2/15/92 and 2/21/92 Figure 1, and HSV-1-specific antibody was detectable in CSF Table 1. EBV-specific DNA was not detectable by PCR in CSF. Additional diagnostic studies that were normal included MRI, EEG, and internal carotid angiography. He received a 10-day course of IV acyclovir (350 mg every 8 hours) without clear improvement. In April 1992, he underwent a lumboperitoneal shunt because of persisting elevation in CSF pressures with associated chronic generalized headaches. The lumboperitoneal shunt malfunctioned and was replaced with a ventriculoperitoneal shunt in May 1993. At that time, a right frontal lobe brain biopsy was performed. The biopsy tissue appeared normal, and there was no evidence of inflammation or viral inclusions. HSV DNA could not be amplified from CSF obtained by lumbar punctures on 5/13/93 and 9/16/93, nor from CSF obtained by cisternal puncture on 2/22/94 Figure 1. CSF HSV-1-specific antibody was absent in CSF on 5/13/93 but detectable at low levels on 9/16/93 and 2/22/94. CSF cultures on 6/24/92, 5/13/93, 9/16/93, and 2/22/94 remained negative for HSV Table 1. He continues to have episodic relapses consisting of headache, mental confusion, and exacerbation of his preexisting motor and sensory abnormalities, but has not developed new signs or symptoms. His symptomatic exacerbations have not responded to further treatment with acyclovir or a course of intravenous foscarnet. He has been gradually tapered off prednisone without obvious deterioration in his condition. His current neurologic examination is largely unchanged from that in February 1992; he continues to have impairment of upward gaze and a mild left hemiparesis, but no significant hyperreflexia and flexor plantar responses. He has mild multimodality sensory impairment in the left upper extremity and mild left-sided dysmetria.

Discussion.

In some patients with HSV encephalitis, ^[1] pathologic injury is maximal within the brainstem rather than the cerebrum. ^{[10-17,20-22]} The diagnosis of HSV BSE, like that of classic encephalitis, has been hampered by the failure of conventional virologic tests on easily accessible clinical material to establish the HSV etiology. Until recently, definitive antemortem diagnosis of HSV CNS infection has required brain tissue, which is not generally feasible in cases of suspected BSE. PCR amplification of HSV DNA from the CSF provides a sensitive and specific method for the diagnosis of herpes simplex encephalitis ^[2-5] and has facilitated diagnosis of other forms of HSV CNS infection. ^[6-9] In this article, we describe the use of PCR amplification of HSV DNA from CSF to establish the diagnosis of HSV BSE.

This patient presented with symptoms suggesting multifocal involvement of the brainstem from the midbrain (upward gaze palsy) to the medulla (facial numbness). Additional findings are consistent with involvement of corticospinal, spinothalamic, lemniscal, and brainstem-cerebellar pathways. Confirmation of the role of HSV in this syndrome is based on amplification of HSV DNA from CSF. Many studies indicate that in immunocompetent patients amplification of HSV DNA from CSF only occurs as a consequence of HSV CNS infection, and HSV DNA is not detected with other neurologic diseases or non-HSV-related CNS infections (see reference 23 for review). Further support for the etiologic role of HSV in this patient comes from CSF antibody studies. The presence of a rising titer of HSV antibody in CSF suggests that either there was intrathecal synthesis of antibody or breakdown of the blood-CSF barrier with transudation of serum HSV antibody into CSF. The absence of detectable CSF antibody against VZV, despite the presence of VZV serum antibody, indicates that the CSF HSV antibody resulted from intrathecal synthesis of antibody rather than from transudation. ^[24-26]

The mechanism by which HSV might produce recurrent BSE remains unclear. The fact that the patient had recurrent herpes labialis indicates that he had HSV latent in the trigeminal ganglia. His early neurologic recurrences may have reflected episodic reactivation of HSV latent in trigeminal ganglia, with subsequent axoplasmic transport of virus centrally into the brainstem rather than peripherally to cutaneous sites. The factors that terminated these initial episodes of BSE are unclear, although the induction of an intrathecal antibody response could conceivably have played a role (see reference 27 for review). He has subsequently had continued, although milder, exacerbations of his symptoms despite the absence of evidence on CSF PCR of HSV reactivation. Possibly, a process other than recurrent viral infection may have contributed to fluctuations in his symptomatology. This would explain why acyclovir was beneficial for his earlier neurologic recurrences but has not been useful in either treating or preventing later exacerbations. The mechanism for these later exacerbations remains speculative. Patients with HSV encephalitis treated with acyclovir may experience relapses. Suggested mechanisms have included persisting infection, new reactivation of virus, and postinfectious encephalomyelitis. ^[28,29] Persisting infection and reactivated infection often respond to renewed antiviral therapy, whereas postinfectious immune-mediated disease does not. We cannot exclude the possibility that this patient's later exacerbations fall into this category, although the absence of white matter disease on repeated MRI studies makes this unlikely. Alternatively, his symptomatic fluctuations may be similar to the periodic worsening of preexisting signs and symptoms that commonly occurs in patients with multiple sclerosis in the absence of new demyelination. ^[30] These attacks may be due to transversely spreading ephaptic activation of axons within previously damaged regions of the CNS.

There is a paucity of confirmed cases of HSV BSE having detailed clinical descriptions. Nonetheless, a number of features commonly occur among the cases reported to date. Neuro-ophthalmologic abnormalities, including ptosis, ^[10,15] anisocoria, ^{[10,11,20,22]} nystagmus, ^{[10,11,14,16,17]} ophthalmoplegia, ^{[10,11,15,20]} and gaze palsies ^[10,20] (our case), are frequently encountered. Involvement of the trigeminal nerve nuclei may produce weakness of masticatory muscles or facial numbness ^{[10,11,14,17,20]} (our case). Some patients have had peripheral facial weakness ^{[10-12,15,17,20]} or findings suggesting involvement of cranial nerves IX to XII. ^{[10,11,14,17,20,22]} Weakness, hyperreflexia, and extensor plantar responses reflecting corticospinal tract involvement may occur, ^{[10-12,14,17,20,22]} and in severe cases patients may become quadriparetic. Our patient had a mild hemiparesis involving the left face, arm, and leg during one attack. Incoordination and ataxia may occur, ^{[10,11,14-16]} presumably as a result of involvement of brainstem-cerebellar pathways. Our patient also had episodic incoordination of the left arm and leg.

The literature suggests that the majority of cases of HSV BSE have been fatal. However, this probably reflects the fact that unequivocal diagnosis previously required the use of postmortem tissue. It is likely that nonfatal cases have been under-reported due to the difficulty in confirming the diagnosis. As illustrated by this report, PCR may facilitate antemortem diagnosis of BSE, which in turn may lead to more accurate definition of the frequency and clinical spectrum of BSE.