Original Articles

Indian Pediatrics 1999; 36: 1097-1106

Investigation of an epidemic of Reye's syndrome in northern region of india 

D. Ghosh, D. Dhadwal*, A. Aggarwal*, S. Mitra⁺, S.K. Garg**, R. Kumar* and   B. Kaur*

From the Departments of Pediatrics, Community Medicine* and Pharmacology**, Post Graduate Institute of Medical Education and Research, Chandigarh 160 012, India and Department of Pediatrics⁺, Government Medical College, Chandigarh, India.
Reprint requests: Dr. D. Ghosh, Assistant Professor, Department of Pediatrics, PGIMER, Chandigarh 160 012, India. E-mail: medinist@pgi.chd.nic.in.
Manuscript received: March 4, 1999; Initial review completed: May 5, 1999; Revision accepted: May 19, 1999.

Objective: To determine the extent, epidemiological and clinical features of an epidemic of non-inflammatory encephalopathy in northern region of India. Design: Surveillance of referred cases having unconsciousness after a short bout of fever during October and November 1997. Case control study in 7 most affected villages. Methods: Active case finding was done to assess the extent and severity of the epidemic by interviewing health professionals and by reviewing mortality records in 10 districs of Haryana, Punjab and Chandigarh. A house to house survey was conducted in seven most affected villages. A case was defined as any child of less than 15 years of age, who had prodromal fever followed by vomiting and unconsciousness with subsequent recovery or death. Two age and sex matched controls who had fever without unconsciousness were taken for each case, one from nearby house and another staying furthest from the affected house. These groups were compared for various epidemiologic factors, clinical features and treatment pattern. Residual medicines used by affected patients were tested for presence of salicylate. Local village practitioners were interviewed for their knowledge and attitude towards use of aspirin in a febrile child. Results: Information regarding 129 affected children (M: F=1 : 1) could be obtained. Age ranged between 1 to 12 years (mean 5.8 years). Most were from rural or semi-suburban areas. Attack rate was 5.4/1000 and case fatality rate was 72%. Multiple sibs were affected in 9.3%. History of fever was reported by 83%, vomiting preceding unconsciousness by 83% and abnormal behavior by 65%. Abnormal posturing was reported in 55%. Seventeen (61%) of 28 samples had IgM antibodies in serum/CSF against measles. Twelve (36%) of 33 serum samples tested positive for Varicella zoster virus. None gave history of aspirin intake and 10 samples of residual drugs did not contain salicylate. However, 6 out of 19 blood samples taken from affected patients contained salicylate. Environmental factors were in favor of Japanese encephalitis (JE) but brain biopsy and serology disproved it. Based on earlier report of JE from this area, the cases in present epidemic were being reported as JE before this study was undertaken. Intensive fogging with malathion was being undertaken as antimosquito measure, specially around the affected houses. Local village practitioners (n=37) were unaware of contraindications of aspirin in a febrile child. Conclusion: Measles and varicella zoster emerged as the probable etiologies for the viral prodrome precipitating these cases of Reye's syndrome. Aspirin might have a contributory role. Malathion is another putative cofactor.

Key words: Encephalopathy, Epidemic, Japanese encephalitis, Measles, Salicylates, Varicella zoster.

From October 1997 onwards, there was a spurt of children with encephalopathy being admitted in pediatric emergency service of Postgraduate Institute of Medical Education and Research, Chandigarh. These children hailed from adjoining districts of Haryana, Punjab and Uttar Pradesh. The initial 2 weeks witnessed around 20 such cases. They presented with prodromal fever, vomiting, abnormal behavior progressing to rapidly oncoming coma leading to death in majority despite intensive care in our hospital. These cases were afebrile at admission, though illness was reported to have started with fever. Thereafter vomiting and agitated behavior progressing rapidly to coma and death occurred in about 12 to 48 hours. There was more than three times rise in aminotransferases with normal serum bilirubin, abnormal coagulogram and increased serum free fatty acid. CT scan of head showed diffuse cerebral edema. Post-mortem liver biopsy showed diffuse micro-vesicular fatty infiltration in absence of inflammation or necrosis, and brain biopsy (postmortem) showed features of cerebral edema or hypoxia/ischemia without any evidence of meningitis or encephalitis. Overall picture of these cases in initial 2 weeks was suggestive of Reye's syndrome. Serum for salicylate was positive in 6 out of 19 cases though none gave history of salicylate use and none of prescriptions contained it. A field investigation was subsequently conducted to determine: (a) the true extent of the disease as cases referred to our hospital might represent only a fraction of whole problem; (b) the etiopathogenic factor or co-factors of the disease, (c) the clinical features; and (d) the natural history of the disease.

Subjects and Methods

Field investigations were carried out in five districs of Haryana (Ambala, Panchkula, Kurukshetra, Kaithal and Karnal) besides Chandigarh by a team comprising of a pediatri-cian and a community physician. Operational definition for case finding included children below 15 years of age, who had unconscious-ness with or without fever and vomiting after September 1997.

Epidemiologic case sheets were prepared in local language (Hindi) based on the experience of clinical features of cases coming to Pediatric Emergency of PGI, Chandigarh. These case sheets were given to medical officers and health workers of various primary health centers. These health professionals were explained about the signs and symptoms of the disease and trained in recording data on the epidemiologic case sheets. Active case finding operation included addressing and interacting with the government and non-government health organizations, meeting the private practitioners and sharing their experience, review of mortality data in civil hospitals, admission data from PGIMER and Government Medical College, Chandigarh. Data were also collected from the civil hospitals of the area by directly examining the admitted patients, and visit to the house of children who had died and enquiring about other cases in their locality.

A house to house survey was conducted in seven most affected villages using a question-naire. The records included total number of children in the family below 15 years of age and any child with fever having upper respiratory infection, diarrhea, skin rash, jaundice, bleeding, vomiting, abnormal behavior, posturing or unconsciousness. Name of medicine given for treatment of fever was also recorded. The prescription and drug stock at home was checked. The detailed epidemiologic case sheets mentioned above were filled for cases. Samples of unlabelled drug stock obtained from household were assayed by HPLC method in Pharmacology Department of PGI, Chandigarh for presence of salicylate.

After obtaining basic data regarding the extent of the problem, a case-control study was done in the most affected villages where house-to-house survey was conducted. A case was defined as a child under 15 years of age who had fever followed by vomiting and uncon-sciousness with subsequent death or recovery during the months of October and November 1997. Age and sex matched controls were selected from the same village who had fever but did not become unconscious during the same period. Two controls were chosen for each case, one from near the affected house and the other from the furthest end of the village.

Private medical practitioners were inter-viewed regarding the drugs they use for treatment of fever in children, whether they used aspirin in children; if not, then why not. They were also asked whether other practi-tioners in their locality use aspirin in a febrile child. Their stocks were also checked for aspirin. Importance of early diagnosis with high index of suspicion, prompt management even with minimal facility and referral guidelines were given to them. Role of aspirin in causation of Reye's syndrome was highlighted and all the health professionals were educated about avoidance of aspirin in a febrile child.

Health education intervention consisted of stoppage of aspirin use in febrile children, prompt diagnosis and management of Reye's syndrome by use of mannitol and frusemide even from early stage, giving 10% glucose containing fluid and vitamin K injections to affected children. Trend of the disease was assessed by comparing the frequency and  severity of disease in early and late stage of the epidemic.

Results

A total of 129 (79 prospective and 50 from records, report by parents/doctor) children with encephalopathy were identified. Out of them, 72% died. The highest number of cases were from Kurukshetra (n=49) followed by Karnal (n=23), Ambala (n=12), Kaithal (n=8), Yamunanagar (n=6), Panchkula (n=4) and adjoining areas of Uttar Pradesh (n=8), Punjab (n=14) and Chandigarh (n=5) (Fig. 1).

Fig. 1. Spot map showing the distribution of cases in Northern India.

Age of the affected children varied from 1 year to 12 years (mean age 5.8±2.5 years). Overall male - female ratio was 1:1. All these patients came from rural or suburban areas where mosquitoes and pigs are widely prevalent. The clinical features of cases are summarized in Table I.

Table I^__ Clinical and Environmental Features of Encephalopathy Epidemic in Northern India, 1997

House to House Survey

This was concluded during the last week of October and whole of November by the first two authors (DG, DD) and the last author (BK). A total population of 13,260 was covered by the house-to-house survey in 7 most affected villages of Haryana. In these villages, children below 15 years were 3163. Village-wise distribution of population surveyed is shown in Table II. Out of 3163 children, 312 (9.9%) had fever, 401 (12.7%) had symptoms of upper respiratory tract infections, 61 (1.9%) were reported to have vomiting, and 52 (1.6%) had diarrhea in last one month.

Table II-Coverage of House-to-House Survey

Case Control Study

TABLE III-Comparison of Cases and Controls

Figure in parantheses are percentages.
*Fever followed by unconsciousness; **Fever not followed by unconsciousness

Comparison of various epidemiological and clinical features of cases and controls are shown in Table III. History of URI was significantly more common among controls whereas history of abnormal behavior, seizures and abnormal posture was significantly more common in cases. More cases came from poorer socio-economic status families living in kutcha houses (59 vs 35%). Proximity to pigs was more in cases when compared with control, but the difference was not statistically significant (Table III). Treatment history in case control study was found to be similar (Table IV).

Table IV^__ Treatment History in Cases and Controls

Interview of Local Practitioners

Thirty-seven local private medical practitioners were interviewed from affected villages visited for case finding. When asked about the drugs used by them to treat fever in children, 10 mentioned name of aspirin alongwith other medications like paracetamol, septran, amoxycillin, etc. When asked directly whether they use aspirin in febrile child, two more persons agreed. Rest of the 25 persons were asked about the reason for not using aspirin; none could reply satisfactorily. Reply varied from "do not know", "strong drug", "allergic reaction", "can cause collapse", "not needed". When asked whether other practi-tioners use it, none answered as "no", 23 were noncommittal, 14 answered positively. On stock checking, 17 out of 37 were found to keep disprin tablet in their stock.

Ninety two cases were reported in October 1997. Of them 73 died, 8 survived and follow up of 11 was not available. Health measures mainly towards primary prevention in the form of stoppage of aspirin use in a febrile child, and secondary prevention in form of sensitising health professional in early diagnosis, prompt management or referral were initiated after the 3rd week of October 1997. The number of cases reported in November 1997 was 37, of whom 16 died and 21 survived. Case fatality rate seems to be higher in October and declined in November 1997.

Discussion

Out of the 129 cases identified by active case finding, 51 had been admitted in the Postgraduate Institute of Medical Education and  Research, Chandigarh. Following findings suggestive of Reye's syndrome were observed in these cases(1).

Classic progression after prodromal fever to onset of vomiting (stage I) to stage of agitated confusion (stage II) and finally altered sensorium with decortication (stage III), decerebration (stage IV) and flaccidity (stage V) and death. Recovery, if occurred, was from any stage (other than V)(1-5). Abnormal coagulogram was observed in 80% with normal total leukocyte and platelet count(6). Malaria parasite was negative in smears. Elevated liver enzymes to >3 times normal (91%) with normal serum bilirubin level was seen in majority (65%)(7). Serum free fatty acid was elevated in 24/24 (100%) cases(8). Hypoglycemia was seen in 26% cases(9). Cerebrospinal fluid was essentially normal except mild rise in lymphocytes in 3 and protein in 4 cases which might be due to prolonged convulsion(10). Contrast enhanced CT of head uniformly showed features of diffuse cerebral edema without any focal lesion(11). Electroencephalo-graphy showed diffuse slowing (7/7 cases) consistent with encephalopathy(12). Post-mortem biopsy/autopsy showed diffuse micro-vesicular steatosis of liver in 19/19 cases (100%)(13). Same changes were seen in renal tubules (5/5 cases). Brain pathology (n=18) showed features of cerebral edema, ischemic/hypoxic neurons in absence of any meningitis or encephalitis(14). These features, taken together, are diagnostic of Reye's syn-drome(4,5,10). Viral studies for Japanese encephalitis, herpes simplex, dengue, West Nile, influenza A and B were negative. Seventeen of 28 serum and/or CSF samples tested seropositive for measles and 12 out of 33 serum tested positive for Varicella zoster. Serum of 6 children out of 19 affected, however, was containing salicylate though history of aspirin intake was not available.

Age group of affected children was relatively lower (mean 5.8±2.5 years) when compared to cases described in western world(15,16). Most studies have found female sex preponderance but in our study they were equal. When patients reaching PGIMER were separately analyzed males outnumbered females (M : F = 3 : 2). The reason may be extra preference given to male children in our society, thereby more of male children were brought to tertiary care hospital.

The present epidemic occurred during early winter which is not very classic of what was found in earlier epidemics in west, where it used to occur during late winter and early spring(10,15). Presence of fever almost universally as a prodromal symptom and turning afebrile when encephalopathy starts, suggest some infection, most commonly viral infection as the priming event but not directly involving organ like brain and liver. This phenomenon goes well with Reye's syndrome. This feature is odd for various other differential diagnoses of epidemic occurrence of un-consciousness like cerebral malaria and viral encephalitis (Japanese B encephalitis). Two viruses (measles and Varicella zoster) emerged as the possible etiological factors. In epidemic form of Reye's syndrome other studies have found influenza A or B as the commonest causative agent(10,15,16). Measles and Vari-cella zoster also figured in previous studies as causative agents(17,18).

Role of aspirin in producing mitochondrial dysfunction in virus primed liver is proved beyond doubt and body of evidence favours aspirin as the most important co-factor for production of Reye's syndrome(19-21). In our study we could not get any history of aspirin use for prodromal fever either in the cases or in the controls. Ten specimens of the drug given for controlling fever, collected from households of children dying from the disease, were analyzed for aspirin. None contained any salicylic acid.   Serum samples taken on admission after variable period from onset of fever detected salicylate in 6 out of 19 tested. Presumably more percentage might have consumed the drug as half life of salicylate is 6 to 8 hours and after a few half lives the serum level may be undetectable.

This prompted us to evaluate the medication behavior of local unqualified private medical practitioners. Many of them use aspirin in a febrile child because of lack of knowledge about its cause-effect relationship with development of Reye's syndrome. This practice is still continuing inspite of ban on its use in a febrile patient imposed by Haryana government because of further worsening of platelet function in dengue fever which was quite rampant in a neighbouring state in the previous year.

Almost all children hailed from rural or semi-suburban areas which goes well with the results of previous studies(22). This epidemic occurred over a wide geographic area covering 10 districts of Haryana, Uttar Pradesh, Punjab and Chandigarh. This area is a rice growing area which is wet most of the times of the year, more so during rainy season. None of the cases was reported from adjacent dry area like Hisar, Rohtak, Kalka, etc. There are widespread mosquito breeding sites over these wet lands mostly after rainy season, i.e., September. Pigs are also widely prevalent in the affected area. The environmental factors remind us of Japanese encephalitis as the probable diag-nosis(23). In 1992, there was an epidemic of unconsciousness mainly centering around Kurukshetra involving almost same geographic area. That epidemic was diagnosed to be probably due to Japanese encephalitis because of seropositivity in some patients(24).

Japanese encephalitis as the cause for the present epidemic is ruled out because of: (i) affection of only children below 12 years, (ii) family clustering of cases seen in 12 cases, (iii) absence of fever after onset of altered sensorium, (iv) deranged liver function with elevated serum aminotransferases and abnormal coagulogram, (v) postmorterm brain biopsy or autopsy did not show any evidence of inflammation or encephalitis and (vi) serology for JE virus (blood and CSF) was negative. This area is however under close surveillance for Japanese encephalitis. Since 1992 epidemic of JE, all cases of fever and unconsciousness are being reported as Japanese encephalitis by the health workers. Fogging is done in this area using malathion as anti-mosquito measures. When any death is reported from a village the spraying becomes even more intense, more so around the kutcha houses of poor people with poor hygiene living in company of pigs. In literature there is report of insecticides causing Reye's syndrome both in experimental animal and human situation(8). Insecticides are well known agent to alter mitochondrial permeability barrier and thereby act as a co-factor in causation of Reye's syndrome(25). However, this is an epidemiologic postulation, yet to be proved in our situation. Alternate anti-mosquito measures might be of great use if the above mentioned cause effect relationship is established beyond doubt.

In the case control study, a few important points emerged. The attack rate in this epidemic was high(54/1000 childhood population below 15 years). CSF and/or serum positivity against measles and serum positivity against Varicella zoster in some cases may indicate subclinical infection as the prodromal illness for develop-ment of Reye's syndrome. Only seropositivity without any clinical feature of measles or zoster or chickenpox is an odd finding.

Exclusion of viral hepatitis leading to hepatic encephalopathy was done by the presence of nonicteric hepatopathy without any inflammation or necrosis in liver other than  some changes of shock in a few. Cerebral malaria was ruled out as this was a situation of afebrile encephalopathy whereas fever will peak with onset of unconsciousness in cerebral malaria. Malaria parasite was not seen in any of peripheral blood film. Dengue encephalitis, though can occur, is seen in upto 5% cases. Normal platelet count, normal hematocrit, absence of brain inflammation or encephalitis rules that out.

Though 2 months are too short a period to conclude about the trend of the disease, there was a distinct reduction in frequency and severity of the cases in 2nd month. This may be attributed to the primary preventive measures to avoid aspirin use in a febrile child or to secondary preventive measures in form of improvement in early diagnosis and manage-ment of clinical cases. However, the natural history of illness may also explain this improvement.

To conclude, whereas all over the globe there is significant reduction of cases of Reye's syndrome, we encountered first ever epidemic in a devastating form in Northern part of India. Use of aspirin in a febrile child might have precipitated the illness following a viral fever (measles or Varicella zoster). This epidemic had high case fatality and progressed rapidly compared to previous epidemics. Insecticide spray of malathion may be a co-factor for development of Reye's syndrome.

We finally recommend imposition of strict ban over use of aspirin in a febrile child. There is a need for health education of poorly qualified local practitioners to avoid use of nonprescript medicines. We need to educate public not to promote self medication. Alternative anti-mosquito measure, other than technical malathion and better prevention of measles and Varicella zoster may potentially help in prevention of such an epidemic.

Acknowledgement

We are grateful to Prof. B.K. Sharma, Director, Prof. L. Kumar, Head and Dr. Sunit Singhi, Additional Professor, Department of Pediatrics, Postgraduate Institute of Medical Education and Research, Chandigarh, to allow us to conduct this study. We also acknowledge the help provided by Dr. P.L. Jindal, Director General Health Services, Haryana and his staff. The contributon by Department of Virology, PGI and National Institute of Virology, Pune is deeply acknowledged.

References

1. Ghosh D, Singhi S, Radotra BD, Ratho RK, Garg SK. Epidemic non-inflammatory hepato-encephalopathy in children: Reye's syndrome resurgence? Neurology India 1998; 46: S34-S35.

2. Lovejoy FH, Smith AL, Bresnan MJ, Wood JN, Victor DI, Adams PC. Clinical staging in Reye's syndrome. Am J Dis Child 1974; 128: 36-41.

3. Rey RDK, Morgan G, Bober J. Encephalopathy and fatty degeneration of the viscera, a disease entity in childhood. Lancet 1963; ii: 749-752.

4. Gauthier M, Guay J, Lacroix J, Lortie A. Reye's syndrome: A reappraisal of diagnosis in 49 presumptive cases. Am J Dis Child 1989; 143: 1181-1185.

5. Hardie RM, Newton LH, Bruce JC, Glasgow JFT, Mowat AP, et al. The changing clinical pattern of Reye's syndrome 1982-1990. Arch Dis Child 1996; 74: 400-405.

6. Hurwitz ES, Nelson DB, Davis C, Morens D, Schonberger LB. National surveillance for Reye's syndrome: A five year review. Pediatrics 1982; 70: 895-900.

7. Ludvigson P, Grover WV, Brown LW. The fallacy of varicella hepatitis: Evidence for Reye's syndrome. Ann Neurol 1981; 10: 300.

8. Davis Le. Reye's syndrome. In: Handbook of Clinical Neurology, Vol. 56: Viral Disease, Ed. McKendall RR. Amsterdam, Elsevier Science Publishers, 1989; pp 149-177.

9. Glasgow AM. Metabolic abnormalities in Reye's  syndrome. J Natl Reye's Syndrome Foundation 1983; 4: 24-31.

10. Barrett MJ, Hurwitz ES, Schonberger LB. Changing epidemiology of Reye's syndrome in the United States. Pediatrics 1986; 77: 598-602.

11. Russel JE, Zimmerman RD, Leeds NE, French J. Reye's syndrome: Computed tomographic documentation of disordered intracerebral structure. J Comput Assist Tomogr 1979; 3: 217-220.

12. Aoki Y, Lombroso CT. Prognostic value of electroencephalography in Reye's syndrome. Neurology 1973; 23: 333-343.

13. Chang LW, Gilbert EF, Tanner W, Moffat HL. Reye's syndrome. Arch Pathol 1973; 96: 127-132.

14. Manz HJ, Colon AR. Neuropathology, pathogenesis and neuropsychiatric sequelae of Reye's syndrome. J Neuro Sci 1982; 53: 377-395.

15. Sullivan-Bolyai JZ, Corey L. Epidemiology of Reye's syndrome. Epidemiol Rev 1981; 3: 1-26.

16. Rogers MF, Schonberger LB, Hurwitz ES, Rowley DL. National Reye's syndrome surveillance, 1982. Pediatrics 1985; 75: 260-264.

17. Noble GR, Corey L, Rubin RJ. Virologic components of Reye's syndrome. In: Reye's Syndrome, Ed. Palloack JD. New York, Grune and Stratton, 1975; pp 189-197.

18. Sullivan-Bolyai JZ, Marks JS, Johnson D, Nelson DB. Reye's syndrome in Ohio (1973-1977). Am J Epidemiol 1980; 112: 629-638.

19. Arrowsmith JB, Kennedy DL, Kuritsky JN, Faich GA. National patterns of aspirin use and Reye's syndrome reporting, United States, 1980 to 1985. Pediatrics 1987; 79: 858-863.

20. Porter JDH, Robinson PH, Glasgow JFT, Banks JH, Hall SM. Trends in the incidence of Reye's syndrome and the use of aspirin. Arch Dis Child 1990; 65: 826-829.

21. Khan AS, Kent J, Schonberger LB. Aspirin and Reye's syndrome. Lancet 1993; 341: 968.

22. Ruben FL, Streiff EJ, Neal M, Michaels RH Epidemiologic studies of Reye's syndrome: Cases seen in Pittsburgh, October 1973-April 1975. Am J Public Health, 1976; 66: 1096-1098.

23. Kumar R, Misra PK. Japanese encephalitis in India. Indian Pediatr 1988; 25: 354-360.

24. Prasad SR, Kuma V, Marwaha RK, Batra KL, Ratho RK, Pal SR. An epidemic of encephalitis  in Haryana: Serological evidence of Japanese encephalitis in a few patients. Indian Pediatr 1993; 30: 905-910.

25. Trost LC, Lemasters JL. The mitochondrial permeability transition: A new pathophysio-logical mechanism for Reye's syndrome and toxic liver injury. J Pharmacol Exp Therap 1996; 278: 1000-1005.