|
|
| |||||||||||||||||||||||||||||||||||||||||||||
Background Between September 1998 and June 1999, there was an outbreak of severe viral encephalitis due to Nipah virus, a newly discovered paramyxovirus, in Malaysia.
Methods We studied the clinical features of the patients with Nipah virus encephalitis who were admitted to a medical center in Kuala Lumpur. The case definition was based on epidemiologic, clinical, cerebrospinal fluid, and neuroimaging findings.
Results Ninety-four patients with Nipah virus infection were seen from February to June 1999 (mean age, 37 years; ratio of male patients to female patients, 4.5 to 1). Ninety-three percent had had direct contact with pigs, usually in the two weeks before the onset of illness, suggesting that there was direct viral transmission from pigs to humans and a short incubation period. The main presenting features were fever, headache, dizziness, and vomiting. Fifty-two patients (55 percent) had a reduced level of consciousness and prominent brain-stem dysfunction. Distinctive clinical signs included segmental myoclonus, areflexia and hypotonia, hypertension, and tachycardia and thus suggest the involvement of the brain stem and the upper cervical spinal cord. The initial cerebrospinal fluid findings were abnormal in 75 percent of patients. Antibodies against Hendra virus were detected in serum or cerebrospinal fluid in 76 percent of 83 patients tested. Thirty patients (32 percent) died after rapid deterioration in their condition. An abnormal doll's-eye reflex and tachycardia were factors associated with a poor prognosis. Death was probably due to severe brain-stem involvement. Neurologic relapse occurred after initially mild disease in three patients. Fifty patients (53 percent) recovered fully, and 14 (15 percent) had persistent neurologic deficits.
Conclusions Nipah virus causes a severe, rapidly progressive encephalitis with a high mortality rate and features that suggest involvement of the brain stem. The infection is associated with recent contact with pigs.
The isolation of a new paramyxovirus, subsequently named Nipah virus, from cerebrospinal fluid specimens from several patients indicated that this was the etiologic agent.3 Preliminary studies of nucleotide sequencing revealed that this virus is closely related to, but not identical to, Hendra virus, which caused disease among horses and affected three patients in Australia.1,6,7
We describe the clinical and laboratory features of patients with Nipah virus infection who were seen at the University of Malaya Medical Center, Kuala Lumpur, one of the two main hospitals to which patients were admitted during the outbreak.
Methods
The patients were treated by a team of neurologists, infectious-disease physicians, and intensive care physicians. Clinical observations and base-line laboratory results were recorded prospectively. Cerebrospinal fluid was examined in nearly all patients, and electroencephalography and magnetic resonance imaging (MRI) of the brain, with a 1.5-tesla scanner (Siemens), were performed in some patients.
Isolation of virus was not attempted in all patients for fear of laboratory-acquired infection. When viral isolation was attempted, syncytial-cell formation on Vero cell cultures (American Type Culture Collection and certified cell line 81) several days after inoculation indicated the presence of the virus.3 Serum and cerebrospinal fluid samples were tested with an IgM-capture enzyme-linked immunosorbent assay (ELISA) and indirect IgG ELISA for antibodies against Hendra virus antigens.1,3 These tests were developed for use and found to be effective during the outbreak, pending the development of antibody tests that used Nipah virus antigens.
Because confirmatory tests (viral isolation or Nipah virus serologic tests) could not be readily performed, a case definition based on epidemiologic and clinical features was adopted. Patients were considered to have Nipah encephalitis if they came from areas known to be involved in the outbreak, had had direct or close contact with pigs or other infected animals, and had evidence of encephalitis. Encephalitis was defined by the presence of one of the following: clinical features (fever, headache, altered sensorium, or focal neurologic signs), abnormal cerebrospinal fluid findings (
6 lymphocytes per cubic millimeter or a protein level of at least 0.45 g per liter in patients under 50 years of age and a level of at least 0.55 g per liter in patients who were 50 years of age or older), or characteristic findings on MRI of the brain. Patients with positive antibodies against Hendra virus antigens but no clinical or laboratory evidence of neurologic involvement were considered to have nonencephalitic Nipah virus infection.
Statistical analysis was performed with the 
2 test or Student's t-test. The prognostic factors for outcome of the infection were analyzed by multivariate logistic regression.
Results
Characteristics of the Patients
Between February 1999 and June 1999, a total of 110 patients were hospitalized with suspected Nipah virus infection. They were mainly from Bukit Pelanduk and neighboring villages in Negri Sembilan State south of Kuala Lumpur. Ninety-one (83 percent) fulfilled the criteria for Nipah encephalitis, and three additional patients had nonencephalitic Nipah infection. The mean age of the 94 patients with Nipah infection was 37 years (range, 13 to 68). The ratio of male patients to female patients was 4.5 to 1. The majority (83 percent) were ethnic Chinese, 14 percent were ethnic Indians, and 3 percent were of other races or ethnic groups. Eighty-seven (93 percent) were pig farmers or had occupations that involved direct contact with pigs, including working in an abattoir, selling animal vaccines, repairing pig cages, and involvement in pig-culling operations during the outbreak. Of these patients, 41 percent reported that the animals with which they had recently had contact had died after contracting an unusual respiratory tract illness. The period between the patient's last contact with pigs and the onset of illness ranged from several days to two months, and was two weeks or less in 92 percent of patients. Two patients had had direct contact with infected dogs (the dogs died of suspicious causes just before the patients became ill). The remaining patients (5 percent) lived close to but were not in direct bodily contact with pigs. More than half (56 percent) had affected family members, and 74 percent had been immunized against Japanese encephalitis virus.
Acute Clinical Features
Table 1 lists the main features at presentation. The mean duration of fever before admission was 3.5 days (range, 1 to 14). The disease predominantly affected the nervous system. The neurologic features are summarized in Table 2. Fifty-two patients (55 percent) had a reduced level of consciousness, as evidenced by a score of less than 15 on the Glasgow Coma Scale; the mean score at nadir (defined as the worst conscious level or the need for mechanical ventilation) was 7.5. The mean duration of the illness from the onset of symptoms to the nadir was 6.9 days (range, 3 to 31). Patients with a reduced level of consciousness had prominent signs of brain-stem dysfunction, including abnormal doll's-eye reflex, pinpoint pupils with variable reactivity, and prominent vasomotor changes consisting of hypertension and tachycardia, which suggested involvement of the medullary vasomotor center. Seizures occurred in 23 percent of all patients, and all but one of these patients had generalized tonic–clonic seizures; that patient had focal motor seizures with secondary generalization. No patient had status epilepticus. Segmental myoclonus, characterized by focal, rhythmic jerking of the muscles, was present in 32 percent. Eight patients had cerebellar dysfunction, and three had severe postural tremor of both arms.
|
|
All other systems were normal in all patients on admission. There was a history of major medical problems in 12 percent of the patients, including hypertension, diabetes mellitus, gout, ulcerative colitis, and gastritis. However, none of these conditions caused clinically significant morbidity. The complications in severely ill patients included systemic sepsis (24 percent), bleeding from the gastrointestinal tract (5 percent), renal impairment (4 percent), hemothorax from insertion of a central catheter (2 percent), and pulmonary embolism (1 percent). Acute atrial fibrillation developed in two comatose patients. Echocardiography in both patients showed normal myocardial function.
Laboratory Investigations
Thrombocytopenia (defined as a platelet count of less than 140,000 per cubic millimeter) was found in 30 percent of patients and leukopenia (defined as a white-cell count of less than 4000 per cubic millimeter) in 11 percent. Blood urea, creatinine, and electrolyte levels were normal in all patients. Elevated levels of alanine aminotransferase (>65 U per liter) were found in 33 percent of patients, and elevated levels of aspartate aminotransferase (>37 U per liter) were found in 42 percent of patients. In 6 percent of patients, the chest radiographs were abnormal, with increased focal markings over the lung fields and, in one patient, mild basal atelectasis. All of these patients had primarily neurologic features at presentation, and only one had cough.
Cerebrospinal fluid examination was carried out in 92 patients (Table 3). The two patients who declined to undergo lumbar puncture had clinical encephalitis, and one of them died. The results of initial and repeated examinations were abnormal (as indicated by elevated white-cell counts, elevated protein levels, or both) in 75 percent and 77 percent, respectively. There was no correlation between abnormal cerebrospinal fluid findings and the severity of disease. The mean Glasgow Coma Scale score at nadir was 11.3 in those with abnormal findings and 9.5 in those with normal findings (P=0.1).
|
|
Among the 18 patients from whom cultures were obtained in order to isolate Nipah virus, the virus was isolated from cerebrospinal fluid in 5. Among the 26 patients from whom cultures from other body fluids were obtained, the virus was isolated from tracheal secretions in 5, from urine in 4, and from urine, nasal secretions, and tracheal secretions in 1 (Figure 2). Tests for antibodies against Hendra virus were performed in 83 patients. Serum samples were positive in 71 percent of patients, cerebrospinal fluid samples were positive in 31 percent of patients, and either or both types of samples were positive in 76 percent of patients. There were no significant differences in demographic and clinical characteristics between patients with antibodies against Hendra virus and the other patients. Among those who were negative for antibodies against Hendra virus or who were not tested, 61 percent had segmental myoclonus. Furthermore, Nipah virus was isolated from two patients who were seronegative and from three patients in whom serologic tests for antibodies against Hendra virus were not performed.
|
Treatment was supportive. Half the patients required mechanical ventilatory support. Seizures were controlled with intravenous phenytoin. Because histopathological studies had shown vasculitis-induced thrombosis,3 aspirin and pentoxifylline were used empirically in 85 percent and 84 percent of patients, respectively. Ribavirin, an antiviral nucleoside analogue, was administered to 78 percent of patients either orally or, in severely ill patients, intravenously.
Thirty patients (32 percent) died, 50 patients (53 percent) recovered fully, and 14 patients (15 percent) had residual neurologic deficits. The mean time from the onset of illness to death was 10.3 days (range, 5 to 29). The direct cause of death in all but two patients was thought to be Nipah encephalitis; one of these two patients died of a massive intracerebral hemorrhage on the 29th day of illness after recovery from coma, and the other died of severe sepsis. All patients with normal levels of consciousness recovered fully after a mean duration of illness of 14.1 days (range, 6 to 24), whereas only eight of the patients (15 percent) with reduced levels of consciousness did so.
Of the 14 patients with residual deficits, 5 remained in a vegetative state. Two patients had residual cognitive impairment and were dependent on caregivers. Three patients had mild cognitive disabilities, and two had mild cerebellar disabilities. Two other patients had residual deficits after having a relapse of the disease.
Factors that appeared to affect the survival of patients are summarized in Table 4. As compared with the patients who survived, the patients who died were older and had more severe brain-stem involvement, as evidenced by the presence of a reduced level of consciousness, vomiting, abnormal doll's-eye reflex, abnormal pupils, hypertension, and tachycardia during the course of the illness. Other poor prognostic factors were the presence of segmental myoclonus, seizures, areflexia, more greatly elevated hepatic aminotransferase levels at admission, and lower platelet counts at admission. There appeared to be no significant difference in outcome with ribavirin treatment. Multivariate logistic-regression analysis showed that an abnormal doll's-eye reflex (P=0.025) and tachycardia (P<0.001) were significantly associated with mortality. The estimated probability of mortality in the presence of both factors was 0.84, whereas it was 0.02 in the absence of both factors.
|
Four patients had late neurologic dysfunction. Three patients had a relapse 13 to 39 days after an initial mild illness. Another patient was seropositive for antibodies against Hendra virus but remained asymptomatic until he became ill 11 weeks later. He subsequently died. One had an isolated nuclear third-nerve palsy, from which she gradually recovered. Three others had altered consciousness, and two of these patients had seizures. Dysphasia and segmental myoclonus with ataxia occurred in one patient each. In three patients, the cerebrospinal fluid examination results were abnormal, with lymphocytic pleocytosis and elevated protein levels; the virus could not be isolated from the cerebrospinal fluid. In patients with relapses or disease of delayed onset, diffuse confluent involvement of the cortical gray matter was found on MRI (Figure 1B).
Histopathological Findings
Five autopsies were performed. The findings from three patients who died within two weeks after the onset of illness have been described previously.3 These patients had mainly disseminated microinfarctions of the central nervous system as a result of vasculitis-induced thrombosis. The remaining two patients died more than a month after infection. One of these patients had massive intracerebral and intraventricular hemorrhage with severely increased intracranial pressure. Residual parenchymal inflammation and vasculitis were still evident at autopsy. The hemorrhage was attributed to vascular fragility resulting from vasculitis. The other patient had delayed-onset disease with encephalitis but no evidence of vasculitis or microinfarction.
Discussion
The outbreak of Nipah encephalitis was one of the most severe outbreaks of any disease in Malaysia. Over 200 people were infected, and the outbreak led to the disruption of the pig-farming industry. Most pig farms were small family farms worked by husbands and wives and sometimes their children. The farmers were mainly ethnic Chinese, and the Malay Muslim residents of the villages did not handle pigs. The fact that most patients were Chinese men suggested that the virus was transmitted by direct contact with pigs or their secretions, rather than by mosquitoes, as in the case of Japanese encephalitis.8 The unusual sudden death of the pigs suggested that the animals were infected with Nipah virus. The infection seems subsequently to have spread from pigs to humans, possibly by contact with infected secretions.1 Since some patients had no direct contact with pigs, the virus may be spread by respiratory droplets. Other animals that appear to be susceptible to the disease, such as dogs, seem to be able to transmit the disease to humans as well. The clustering of cases among family members suggested a high attack rate. Human-to-human transmission has not been documented.1,2
Nipah virus appears to have a short incubation period, during which it spreads systematically, as evidenced by isolation of the virus from urine and throat and nasal secretions. The finding of abnormal chest radiographs in several patients indicates possible pulmonary involvement. Our patients had no substantial respiratory symptoms, although pneumonia was reported in three patients with Nipah virus infection from Singapore.5 Vasculitis has been identified in the lung and several organs other than the brain in patients with Nipah virus infection,3 and it is possible that vasculitis may become severe enough to cause clinical symptoms.
The mental obtundation and neurologic signs indicate that Nipah virus probably has a predilection for the central nervous system. The disease appeared to be nonencephalitic in only a few patients. Meningism was relatively uncommon, suggesting that primary meningitis was not important. The disease was thought to be due primarily to inflammation of the small blood vessels, with thrombosis and microinfarction, although neuronal invasion also seems to occur.3 The mental changes appeared to be due to diffuse cerebral involvement, which was also indicated by diffuse electroencephalographic and MRI abnormalities. However, the prominent brain-stem dysfunction in comatose patients suggests that specific brain-stem involvement may also be important.
Although some neurologic signs were nonspecific and probably reflected widespread vasculitis, other neurologic features appeared to be characteristic and distinctive, including hypertension, tachycardia, segmental myoclonus, and hypotonia and areflexia. These features suggest a predilection of the virus for certain groups of neurons. Segmental myoclonus indicated focal involvement of neurons in the lower brain stem, the upper cervical spinal cord, and the lower spinal cord. We are not aware of any other type of encephalitis characterized by similar segmental myoclonus. Tendon areflexia appeared to be primarily central in origin, because of its generalized nature and diffuse occurrence in the presence of relatively normal muscle power and nerve conduction. However, abnormal late responses in nerve-conduction studies in some patients may suggest spinal-root involvement as well. In patients with persistent flaccid tetraplegia, the absence of upper-motor-neuron signs, a positive response to cortical magnetic stimulation, and a subsequent good motor recovery in most patients argue against the presence of a pyramidal lesion. Recovery of cranial motor function in some patients in whom tetraplegia persisted suggests that other descending tracts from the brain stem may be involved.
Tests for antibodies against Hendra virus were not positive in all patients. This may be because the test was directed against antigens of a different, albeit related, virus. Another reason may have been that some patients died before seroconversion. The facts that the clinical characteristics of seropositive and seronegative patients were similar and that Nipah virus was isolated in some seronegative patients indicate that both seropositive and seronegative patients had the same disease. Furthermore, there is no evidence of the presence of a concurrent epidemic of viral encephalitis due to other agents. With the culling of the pigs, the epidemic rapidly died down, confirming that pig-related Nipah encephalitis was the only epidemic.
The results of initial cerebrospinal fluid examination were abnormal in only 75 percent of patients. Although they were important in confirming encephalitis, the abnormalities were nonspecific and were similar to those of other viral encephalitides. MRI of the brain was more sensitive than cerebrospinal fluid examination, and the multifocal lesions (probably corresponding to areas of microinfarction) appeared to be specific for acute Nipah encephalitis. Electroencephalographic changes were also nonspecific. Unlike MRI findings, the degree of electroencephalographic abnormalities (in particular, the presence of periodic complexes) reflected the severity of neurologic involvement and was related to the outcome.
The disease had a high fatality rate. The main cause of death appeared to be severe neuronal dysfunction, especially in the brain stem. The factors associated with mortality — coma, abnormal doll's-eye reflex, abnormal pupils, hypertension, tachycardia, vomiting, and segmental myoclonus — all indicate severe brain-stem involvement. Higher hepatic-enzyme levels and lower platelet counts on admission in patients who died probably reflected nonspecific systemic changes in very ill patients. Very few survivors had severe residual deficits, in contrast to survivors of other encephalitides, such as herpes simplex encephalitis and Japanese encephalitis.
Neurologic relapse may occur after mild or asymptomatic infection. The failure to culture the virus from cerebrospinal fluid and the different histopathological and MRI changes (affecting mainly cortical gray matter) in patients with relapse suggest that the pathophysiologic process differs from that of acute encephalitis. The course and characteristics of the relapse bear some similarities to those in a patient with Hendra virus infection who died of encephalitis 13 months after recovering from his initial illness.6 The mechanisms of relapse need to be investigated further.
We are indebted to Dr. Thomas G. Ksiazek and the members of the laboratory staff of the Centers for Disease Control and Prevention, Atlanta, for providing Hendra virus serologic tests, and to Dr. John Arokiasamy and Mr. L.H. Ong of the Department of Social and Preventive Medicine, University of Malaya, for help in statistical analysis.
Source Information
From the Divisions of Neurology and Infectious Disease, Department of Medicine (K.J.G., C.T.T., N.K.C., A.K.), and the Departments of Anaesthesiology and Intensive Care (P.S.K.T.), Radiology (S.A.S., B.J.J.A.), Pathology (K.T.W.), and Medical Microbiology (K.B.C., S.K.L.), University of Malaya, Kuala Lumpur, Malaysia.
Address reprint requests to Dr. Tan at the Division of Neurology, Dept. of Medicine, Faculty of Medicine, University of Malaya, 59100 Kuala Lumpur, Malaysia, or at tlip{at}pl.jaring.my.
References
| |||||||||||||||||||||||||||||||||||||||||||||
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2008 Massachusetts Medical Society. All rights reserved. |
