Parkinsonism due to predominant involvement of substantia nigra in Japanese encephalitis

Patients and methods.

Of 52 patients with JE seen over the course of 6 years, five were found on MRI to have selective involvement of substantia nigra. All patients lived in the northeastern districts of Uttar Pradesh (a province of India) near its border with Nepal. JE is known to be endemic in this region.11 This is perhaps a result of yearly postsummer floods in the tributaries of river Ganga (Ganges) and subsequent mosquito breeding in the adjacent rice fields. Of the 52 patients, 43 (including the five described here) were observed between the months of September and February. Diagnosis was confirmed in all five patients based on high immunoglobulin (Ig)M and IgG titers against JE virus as measured by hemagglutination inhibition (HI) test. The HI antibodies were estimated by the technique of Clarke and Casals,12 using a microtiter method.13 The method estimates combined IgG and IgM titers. Generally, fourfold or more than fourfold rise in HI antibody titers in the convalescent phase of illness as compared with the acute phase is considered diagnostic of JE. However, this criterion is somewhat arbitrary, and most studies from this endemic zone have included patients with more than threefold rise in titers as of diagnostic significance.3,14 The convalescent phase study was carried out after about 10 days of the acute phase study. One patient was first examined after the acute phase was over. His blood was also examined for HI antibodies after 10 months of acute illness to look for any delayed decline in titers against JE. Blood and CSF PCR studies were done to exclude herpes simplex and measles encephalitis. CSF was also examined to exclude pyogenic, tubercular, and fungal meningitis. Blood smears were examined on two occasions in all patients to exclude cerebral malaria. Detailed clinical examination was carried out in all patients during acute illness and during convalescence. Three patients were followed for 1 year and two patients for about 2 months.

MRI was performed on a 1.5-T superconducting unit (Siemens; Magnetom, Germany) using a circularly polarized head coil. T2 (repetition time [TR]/echo time [TE]/n = 2200/20/1), T1 (TR 1000/TE 14/n = 3), and magnetization transfer (MT) T1 (TR 1000/TE 14/n = 3)–weighted imaging was done with off-resonance presaturation. Postcontrast MT–T1-weighted imaging was also performed in axial plane in all cases. The imaging was done using 5-mm slice thickness, 0.5-mm interslice gap, and 192 × 256 matrix. For contrast study, Gd-DTPA was injected intravenously in a dose of 0.1 mmol/Kg of body weight. In one patient, T2-weighted sagittal image was also obtained to determine the precise extent of the lesion.

Results.

Clinical, laboratory, and neuroradiologic findings of the five patients are summarized in the table. There were three boys and two girls. Their ages ranged between 7 and 16 years. At the onset of illness, all patients had fever and headache; three of them also had rigors. Two patients had generalized tonic-clonic seizures. Three patients had mildly altered consciousness that developed 3 to 5 days after the onset of fever; two remained conscious throughout their illness. From the beginning of illness or on regaining consciousness, all patients manifested with mutism that lasted several weeks. Later on, when they spoke, they had a very low volume, monotonous but fluent speech. All five patients had decreased movements of the body. Four of them preferred to remain quiet and static; one was irritable and constantly moving around.

Examination findings included restricted extraocular movements in all directions (three patients), upbeating nystagmus (two patients), and opsoclonus (one patient). Each of the five patients had parkinsonian features in different combinations (see the table). Drooling of saliva was observed in three, palmomental reflex in two, and glabellar-tap sign in two patients. All the patients had cogwheel rigidity; four of them had lead pipe rigidity as well. Three patients had tremors at rest, one of them having typical pill rolling tremors.

In all five patients, MRI revealed hyperintensities in the substantia nigra on T2-weighted images that appeared hypointense on T1. MT–T1-weighted images defined the boundary of the lesion as hyperintense with central isointensity. Postcontrast study did not show any abnormal enhancement. All the patients showed normal signals in the rest of the brain parenchyma.

Three patients were followed for 1 year, during which time they recovered completely. Tremors were the earliest to resolve (1 to 4 months) followed by cogwheel rigidity (3 to 8 months). Bradykinesia was the last to resolve (9 to 12 months). In the other two patients, parkinsonian features were recovering when last seen after 2 months of acute illness. A detailed account of selected patients follows.

Patient 1.

A 14-year-old girl developed a fever of 104 °F with severe headache and body ache. There was no seizure, vomiting, or unconsciousness. Four days after the fever, she developed hypophonic speech and restlessness. After 2 days, her relatives noticed stiffness in her body and slowing in all body movements. She moved with short steps and forward stoop. She preferred not to speak most of the time and, on great provocation, spoke in a low volume slurred speech, which remained fluent. On examination, the patient was conscious, well oriented, and cooperative. She had masklike face, drooling of saliva, and parkinsonian gait. She was restless, anxious, and depressed. She preferred to constantly move around whenever left unattended. Her glabellar-tap sign and palmomental reflex were positive. Her movements were extremely slow. Her limbs had cogwheel rigidity that was more marked at wrist, ankle, and forearm supination/pronation. She had normal power, normal deep tendon reflexes, and bilaterally flexor plantar response. Sensory system was normal and cerebellar signs were absent. Her CSF examination revealed normal sugars, 50 mg/dL proteins, and 10 cells/mm³ (all lymphocytes). Serum antibodies against JE virus showed a more than fourfold rise in titers within 12 days. MRI of the brain showed T2 hyperintense lesions confined to substantia nigra in both upper and lower midbrain axial sections. These were hypointense on T1 and of mixed intensity on MT sequence. There was no enhancement on gadolinium contrast injection (figure 1). The patient was followed for 1 year. Her irritability and tendency to move around fully recovered within 1 month. She was still speaking less and in a monotonous low volume voice. Her gait became typically parkinsonian. Depression, drooling of saliva, hypokinesia, and cogwheel rigidity persisted. Examination done between 6 months and 1 year showed gradual improvement in all her parkinsonian features. At 1 year, she had no neurologic deficit.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1. T2 (A), T1 (B), magnetization transfer (MT)–T1 (C), and postcontrast MT–T1 (D) weighted MRI in the axial plane through the midbrain in a patient with Japanese encephalitis show selective lesions in the substantia nigra that are bilateral but asymmetric. The lesions appear hyperintense on T2 (arrows), hypointense on T1, and mixed intensity on MT–T1-weighted images. There is no abnormal enhancement on postcontrast MT–T1-weighted image. The rest of the brain did not show any signal abnormality.

Patient 3.

A 7-year-old girl had headache, vomiting, and low grade fever a month before admission. Two days after the onset of illness, the fever became high, and went up to 105 °F along with rigors. The fever lasted 6 days, and on the fifth and sixth days of fever, she had alteration of consciousness and dystonic posturing of the neck. She developed focal seizures on the 12th day of her illness for which she was put on carbamazepine. During coma, she had stiffness in her entire body and tremulousness on sitting with support. On regaining consciousness after 20 days, she had slowing of her body movements. She did not attempt to comprehend or speak.

Examination carried out 1 month after the onset of illness revealed a conscious but irritable child who did not respond to verbal commands. She had masking of facial expression, drooling of saliva, positive glabellar-tap sign, opsoclonus, generalized rigidity, and cogwheeling at distal joints. Her movements were extremely slow. There were tremors in both hands, which became typical pill rolling when she was made to stand with support. Her CSF showed 70 mg/dL sugars, 60 mg/dL proteins, and 20 cells/mm³ (all lymphocytes). Titers against JE virus were 1:80 and 1:640 in acute and convalescent sera. MRI brain showed hyperintense lesion in the substantia nigra on T2-weighted image (figure 2A), which was hypointense on T1 and of mixed intensity on MT sequence. There was no contrast enhancement (figure 2B). Two months later, she was still mute, with saliva drooling from the side of her mouth. She walked with forward stoop, pill rolling tremors, and bradykinesia.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2. Axial MRI through midbrain in a patient with Japanese encephalitis shows T2 hyperintensities (arrows) in the substantia nigra (A). These lesions were hypointense on T1 and of mixed intensity on magnetization transfer (MT)–T1-weighted images. There was no enhancement on postcontrast MT–T1 image (B). Imaging of the rest of the brain showed no abnormal finding.

Patient 5.

A 16-year-old boy had a high-grade fever of 105 °F along with rigors, headache, and vomiting. The prodromal symptoms subsided in 5 days. Over the next week, he gradually stopped speaking and was not responsive to verbal commands. He had a dazed look and stiffness in his body. His movements also became slow. He remained conscious throughout the illness and did not experience seizures. Examination revealed a mute patient with drooling of saliva and a staring look. His spontaneous eye movements were restricted in all directions. Response to verbal commands was absent. Palmo-mental reflex and upbeating nystagmus were present. He had lead pipe rigidity in all four limbs. Minimal cogwheeling was observed at wrist, ankle, and pronation/supination of forearm. There was no pyramidal tract sign or cerebellar sign. His CSF examination showed 68 mg/dL sugars, 70 mg/dL proteins, and 80 cells/mm³ (all lymphocytes). CSF cultures were negative for pyogenic and tubercular infections. Blood PCR for herpes simplex and measles viruses was negative. Paired sera (1 week apart) showed rising titers for JE (first serum, 1:20; convalescent serum, 1:80). MRI showed symmetric lesions in the midbrain confined mainly to substantia nigra (figure 3). The lesions were hyperintense on T2- and hypointense on T1-weighted images. There was no lesion elsewhere in the brain, including thalamus.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3. T2-weighted axial MRI through the midbrain shows selective hyperintensities (arrow) in the substantia nigra in a patient with Japanese encephalitis.

Examination after 10 days of admission revealed a conscious patient who could comprehend verbal commands. He had masking of face and mutism. His nystagmus and palmo-mental reflex improved. Glabellar-tap habituation was absent. He could move his eyes in all directions. Upgaze and downgaze were normal. He still had drooling of saliva. The most prominent finding was lead pipe rigidity with occasional cogwheeling at distal joints. He walked with forward stoop, generalized bradykinesia, and pill rolling tremors. The tremors responded to anticholinergic agents. Most of the other parkinsonian features responded mildly to levodopa, and the response was better but still incomplete when amantadine was added.

Initial 6 months’ follow-up revealed increase in alertness, improvement in comprehension, and decrease in drooling. Speech became fluent but continued to be monotonous and slurred. Tremors that symptomatically improved with trihexiphenedyl remained improved on gradual withdrawal of the drug. Other parkinsonian features were still present at 6 months and were only partially responsive to levodopa and amantadine. All parkinsonian features gradually improved in the next 6 months, so much so that 1 year from acute illness, he did not require any medication.

Discussion.

The five cases described here had evidence of infection with JE virus. All belonged to an endemic zone in the northeastern region of the Uttar Pradesh province of India.11 This region borders southwestern Nepal, where JE is quite prevalent.15 All the patients had prodromal symptoms and CSF findings compatible with viral encephalitis. Predominant extrapyramidal signs during convalescence were suggestive of JE.16 PCR study for herpes simplex and measles encephalitis was negative in all patients. The diagnosis was confirmed in four of our five patients by HI test, which demonstrated significant rise in titers against JE virus in convalescent serum as compared with acute serum. Although the rise in HI antibody titers in three of our patients was less than the somewhat arbitrary criterion of fourfold rise, it was well within the range observed in different studies on JE from this region.3,14 One patient (Patient 2 in the table) had very high initial titers as his acute phase was already over by the time he was first attended by us. His first as well as second blood sample, withdrawn 10 days apart, showed very high HI titers without showing a rising trend as both of his blood samples represented the convalescent phase. His diagnosis was supported by studying the third blood sample after 10 months of acute illness, which showed considerable fall in the titers. This suggested that his initial rise in HI titers was not related to his basic immunity against JE but due to acute infection with JE virus. None of our patients was vaccinated against JE as the vaccine is not in general use in India.

JE is usually a diffuse meningoencephalitis that involves several areas of the CNS.7,16,17 Autopsy studies have shown involvement of grey as well as white matter of the cerebral hemispheres, thalamus, brainstem, basal ganglia, cerebellum, and spinal cord.4 Reported pathologic features are focal capillary congestion, edema, punched out necrolytic areas, small capillary hemorrhages, perivascular mononuclear infiltration, and necrosis.4,18 These pathologic findings, however, have been observed in severe cases who died of JE. In relatively mild patients who recover, CT and MRI studies have shown widespread lesions distributed in different parts of CNS.17,19 In most of these studies, thalamic involvement has been emphasized as the most important finding. The five cases described here had no thalamic involvement and lesions were confined to substantia nigra in the midbrain region. This suggests that thalamic involvement, although extremely common, is not essential for MR diagnosis of JE. Involvement of midbrain is perhaps equally important and may occasionally be the exclusive site of lesion. MRI evidence of isolated lesions in the substantia nigra, as described here, has never been documented in JE. However, several recent reports include substantia nigra as one of the many sites of involvement in JE. In one autopsy study on human brains, JE viral antigen was immunocytochemically localized to thalamus, hippocampus, substantia nigra, and medulla oblongata.20 A similar study on JE virus–infected mice demonstrated viral antigen mainly in substantia nigra, striatum, cerebral cortex, hippocampus, diencephalon, and other parts of mesencephalon.21 In another human autopsy study, lesions were found in thalamus, substantia nigra, and Ammon’s horn.18 In this study, the lesions in the substantia nigra were mainly in the middle parts of zona compacta. There are anecdotal case reports on JE in which MRI showed lesions in the substantia nigra along with other common sites such as thalamus, hippocampus, globus pallidus, and cerebral white matter.10,22

The current description of JE patients with isolated lesions in the substantia nigra has great significance in view of the good correlation with the clinical syndrome in which parkinsonian features predominated. Such selective lesions have been demonstrated with toxins such as MPTP. However, the current study demonstrates the capability of a virus to do so. The initial clinical course of the five cases described here indicates affection of the brain beyond substantia nigra, as is evident from the presence of one or more than one of the following: alteration of consciousness, opsoclonus, upbeating nystagmus, restriction of eye movements, and seizures. MRI showed lesions in the substantia nigra and nowhere else in the brain. This suggests that significant lesions were confined to this region alone. This is also evident from the follow-up of these patients who showed early improvement in signs other than parkinsonism. Although parkinsonian features also improved within 1 year, it would require a long-term follow-up of these patients or the experimental studies on animals to know whether mild to moderate virulence of the virus can cause partial damage to cells in the substantia nigra, rendering them vulnerable for early abiotrophy at some later stage of life. The significance of this hypothesis lies in the fact that many people develop subclinical infection in the endemic zone, and the possibility of substantia nigra involvement in some of them (and in few others with only prodromal symptoms) cannot be ruled out. This is relevant in view of the fact that all of our patients had a relatively mild clinical course, involvement of only a very small segment of the brain, and complete clinical recovery. They had no or minimal alteration of consciousness, from which they recovered quickly. There were no signs of raised intracranial pressure, no MRI evidence of hemorrhage within the lesions, and no gadolinium enhancement. The vital functions remained normal throughout their illness. Also, the various types of movement disorders described with widespread thalamic and basal ganglia lesions in JE were absent in these patients except for the parkinsonian features.

The parkinsonian syndrome observed in our patients correlates well with MRI evidence of selective involvement of substantia nigra. This observation is also in accordance with experimental evidence from JE virus–infected Fischer rats that suggests decrease in tyrosine hydroxylase-positive neurons in bilateral substantia nigra and resultant decrease in dopamine concentrations in the striatum.9

Several viruses can involve substantia nigra with or without involvement of other parts of the brain. The damage caused by these viruses could be due to release of cytokines from the virus-infected cells or due to direct invasion of substantia nigra by the virus. In one study, 50% of HIV patients with parkinsonism had typical akinetic parkinsonian features that could not be attributed to concurrent use of neuroleptic agents.23 In another study, the number of neuronal cell bodies in pars compacta were 25% lower in AIDS patients despite no direct viral invasion.24 Although these patients had no extrapyramidal syndrome, the subclinical nigral degeneration was implicated in the heightened susceptibility of some AIDS patients to drug-induced parkinsonism.24 Parkinsonism due to lesions in the substantia nigra is also reported with encephalitic illness similar to that of encephalitis lethargica of von Economo’s type.25-27 In one study, fluorodopa PET showed a significant bilateral reduction of tracer accumulation in both putamen similar to that observed in patients with idiopathic PD.27 These findings were suggestive of limited lesions in the dopaminergic neurons of the substantia nigra even though the clinical syndrome also included axial dystonia and stereotyped abnormal movements along with severe parkinsonism.