Ischemic Stroke Secondary to Postinfectious Focal Varicella Vasculitis
Patient #27 presented with the sudden onset of transient inability to talk accompanied by left-sided weakness. These episodes lasted only 30-60 seconds and subsequently resolved (with the exception of the last episode, which left a sustained deficit). The distribution of involvement (face, arm and leg) suggests a subcortical process, though the inability to talk suggests cortical involvement. However, the finding of left-sided weakness associated with a true aphasia would be very unusual unless the person in question was left-handed and the lesion/process involved subcortical as well as cortical structures (in other words, a large stroke). Speech arrest can be seen with sudden neurological deficits in several areas including non-dominant subcortical regions, but the speech arrest is usually transient as in this case.
The child's initial clinical presentation suggested a TIA, embolic stroke, or complex partial seizure with a post-ictal paralysis (though one might also consider a hemiplegic migraine). Stroke in the young is often the result of an underlying metabolic abnormality or overwhelming intercurrent illness. Focal seizures may result from structural abnormalities, stroke, or a parasitic infection such as neurocysticercosis, among others. As the child's illness progressed, and results of the laboratory studies and radiographic images became available, the differential diagnosis was narrowed.
While the child had a history of headaches, there was no previous association with altered consciousness or transient hemiparesis. The duration, quality, and location of the headache pain were also not typical for migraine. The headache pain the child experienced on this occasion was not consistent with migraine, and the frequency of neurological deficits (many brief, transient episodes) was not consistent with hemiplegic migraine episodes. Hemiplegic migraine, therefore, was not strongly considered in this case.
A diagnosis of recurrent partial onset seizures was considered in this case because of the paroxysmal nature and spontaneous resolution of the episodes. The child experienced seven similar episodes within 24 hours and a prolonged video/EEG monitoring study showed right central-occipital sharp waves consistent with a focal epileptogenic process. Furthermore, following IV loading with fosphenytoin, his symptoms temporarily stopped. The left-sided weakness was compatible with a Todd's paralysis. However, a non-contrast CT of the head performed at an outlying institution reportedly showed a large territory right parietal ischemic infarct. MRI performed at our institution showed evidence of a recent infarct in the right head of the caudate. The FLAIR sequences showed a larger area of increased signal intensity that involved not only the right head of the caudate, but also the adjacent lenticular nuclei. These studies supported a vascular etiology, though reactive seizures resulting from an acute stroke remained a possibility.
Initially, an embolic stroke was strongly considered because of the acute onset of symptoms (difficulty thinking of words, left-sided weakness, lack of tonic posturing or clonic movements). The paroxysmal nature of the events, distribution and resolution of symptoms, stereotypy of episodes, and rapid return to baseline function were felt to be inconsistent with a cardioembolic source. Carotid artery embolism (artery to artery embolism) is extremely uncommon in this age group without a predisposing condition, such as collagen-vascular disease or familial hyperlipidemia. In this case, there was no evidence by history or laboratory to support these diagnoses.
Other causes of strokes in childhood include hypercoagulable states (e.g., sickle cell disease, anticardiolipin/antiphospholipid antibodies, low serum protein C or S, antithrombin III deficiency, homocysteinuria, mitochondrial diseases (MELAS), thrombocytosis, and neoplasm). In this case, however, there was no evidence to support any of these diagnoses. Trauma, vascular abnormalities, Moya-Moya disease, postinfectious vasculitis/vasculopathy are additional etiologies for stroke in a child.
Examination of the cerebrospinal fluid showed a pleocytosis (primarily monocytes) with a mildly elevated protein and decreased glucose. Gram stain, AFB stain, and india ink stain were negative, excluding infectious causes. Serum antibody titers for arboviruses and mycoplasma were negative excluding encephalitis and common post-infectious and infectious causes for stroke. Transthoracic echocardiogram did not eliminate the heart as an embolic source, but transesophageal echocardiogram did.
Angiography was performed to examine the cerebral vasculature. There was no evidence of carotid atherosclerotic disease; nor was there evidence of small vessel proliferation to suggest Moya-Moya disease. There was a clot in the M1 segment of the right middle cerebral artery (MCA) at the origin of the lenticulostriate arteries, suggesting either an embolus or focal vasculitis. As no other evidence of an embolic source could be identified, the diagnosis of a focal vasculitis was entertained. The placement of the clot near a vessel bifurcation is typical of many vasculitides.
Vasculitis may be the initial manifestation of a systemic process, such as meningitis, mycoplasma infection, SLE, rheumatoid arthritis, drug abuse (particularly amphetamine abuse), or mixed connective tissue disease. Focal CNS vasculitis, on the other hand, may be the sequela of a preceding viral infection, such as with Epstein-Barr or varicella virus. Furthermore, postinfectious varicella vasculitis has been reported to cause focal angiitis at the M1 segment of the MCA near the origin of the lenticulostriate arteries. In this case, the child had a confirmed bout of chicken pox one month prior to our evaluation.
We diagnosed the child with post varicella focal vasculitis, based on the lack of evidence for an acute intercurrent illness, embolic source, hypercoagulable state, or other underlying disease process and the relative proximity of an uncomplicated bout of chickenpox. Based on the initial transthoracic echocardiogram with poor visualization of the left auricle, we began anticoagulation with heparin until a repeat echocardiogram was performed. When this study showed no evidence of thrombus within the left atrium, anticoagulation was discontinued. Antiplatelet therapy was not instituted. While anticoagulated, and throughout the remainder of his hospitalization, he experienced no further episodes. The left-sided weakness and mild dysarthria slowly resolved. Speech and physical therapy were initiated in the hospital and continued on an outpatient basis.
We saw the patient four months after his hospitalization, at which time his deficits had completely resolved with the exception of mild left weakness in the distal flexors of the upper extremity. He had returned to his normal activities and required no further therapy.
Varicella vasculitis is an uncommon but not unusual sequela of a primary varicella infection. Clinical findings include hemiplegia with dysarthria and/or apahsia. It occurs in previously healthy children with no other identified risk factors for "stoke in the young", but who have had a recent chickenpox infection. The time elapsed from the primary varicella zoster infection to vasculitic symptoms may vary from weeks to months. Since the skin lesions resolve prior to the onset of vasculitic symptoms, varicella vasculitis may be under-recognized as a cause for childhood cerebral infarcts. The exact incidence of this condition is unkown, though it is becoming increasingly recognized as a potential cause for stroke in the young. Based on a series of four patients, Ichiyama et al extrapolates that vasculitis with subsequent stroke may occur in as many as 1:6500 cases of primary varicella infections and be the unrecognized cause of numerous childhood infarcts.
The pathogenesis of varicella vasculitis is unknown. A similar condition may follow infection from herpes zoster ophthalmicus (HZO) and provide some insight. Zoster vesicles erupt over the ophthalmic division of the fifth cranial nerve, often including the orbit. In some cases, an ipsilateral vasculitis with contralateral hemiparesis follows within weeks to months. Central nervous system inoculation in HZO is thought to occur either by hematogenous spread or direct extension of the virus along the ophthalmic division of the fifth cranial nerve to the ipsilateral internal carotid artery. A similar mechanism is postulated for varicella vasculitis, though the mechanism of CNS inoculation is less clear. The proposed vasculitic mechanism concerns an immune attack on viral particles invading local blood vessels rather than an autoimmune process due to molecular mimicry.
The middle cerebral artery (MCA) territory is most often affected by varicella vasculitis. Most case reports note CT evidence of unilateral hypodensities in the basal ganglia and adjacent internal capsule, consistent with occlusion of the origins of the lenticulostriate vessels. Angiography shows focal narrowing of the M1 segment of the MCA contralateral to the side of weakness. Other authors have reported additional "beading" and narrowing of the anterior and posterior cerebral arteries, consistent with more traditional evidence of a vasculitis.
Diagnosis of varicella vasculitis is clinical. Laboratory investigation helps only to include and exclude possible diagnoses; there is no definitive test, aside from brain biopsy. To date, all case reports note a temporal relationship between a preceding clinical varicella zoster infection and the onset of vasculitis. CSF examination is often unrevealing, though a mild pleocytosis is often present. Liu and Holmes, in reviewing the case reports up to 1990, noted that many patients had normal CSF varicella titers. Shuper, Vining, and Freeman, however, noted elevated titers in a patient with oligoclonal banding.
The natural course of the disease is a gradual return to baseline functioning. The disease is self-limited and not thought to warrant overly invasive investigation, such as brain biopsy. However, recurrence of symptoms has been reported in several children, months after their initial presentation. Caekebeke et al. describe the case of a five year old boy whose symptoms returned approximately four months following the first episode. Shuper, Vining, and Freeman report a seven year old boy whose hemiplegia and aphasia recurred seven months later in the same distribution as his earlier deficits. Despite these recurrences, the children again returned to baseline functioning, or had only minimal deficits.
Treatment of varicella vasculitis is not standardized and is primarily anecdotal. Several authors report using intravenous steroids, oral prednisolone, and/or anticoagulation. Others have expressed concern over the use of steroids because it may permit viral resurgence. Clinical improvement is likely independent of treatment, since those who received different treatments had similar good outcomes. Acyclovir may be helpful in preventing further viral production, but its benefit in the resolution of vasculitis is unclear. Anticoagulation may be beneficial early in the course of treatment while embolic sources are investigated. Long term therapy is not, however, indicated.
In summary, varicella vasculitis is characterized by hemiplegia with speech difficulties following varicella zoster infection. The pathogenetic mechanism is unknown, but thought to involve viral invasion of blood vessels and focal vasculitic attack. The preponderance of cases among children results from the greater rate of childhood exposure to varicella zoster. Vasculitic symptoms may occur weeks to months following the primary infection. Despite treatment, the disease appears to be self-limited; residual deficits are minimal to non-existent, though recurrences have been noted several months after the initial insult.
Email comments: