Fatal Outbreak from Ingesting Toxic Plants in Sylhet District

Abstract:

The Institute for Epidemiology, Disease Control and Research, Ministry of Health and Family Welfare, and ICDDR,B investigated a cluster of deaths occurring in Gowainghat and Companyganj upazilas in Sylhet in November 2007. Patients presented with vomiting, restlessness, unconsciousness, and elevated liver enzymes; for patients who died, deathoccurred within hours of symptom onset. In total, 81 patients were identified from 11 villages; 24% (19/81) died. Households who cooked ghagra shak, a local uncultivated plant, were 28.9 times more likely than others to have experienced vomiting and unconsciousness during the outbreak period. Ghagra shak is the local name for Xanthium strumarium which is known to have caused similar illness and death in livestock and children in other countries. Public health messages advising against

eating ghagra shak should be widely disseminated and clusters of patients with vomiting and altered mental status should be promptly reported by physicians to health authorities.

On November 4, 2007 the Director of Institute for Epidemiology, Disease Control and Research (IEDCR) was notified about a cluster of deaths in Sylhet District in northeastern Bangaldesh. A woman and one of her children from a village in Gowainghat upazila presented to Gowainghat Health Complex with unconsciousness following a brief illness of vomiting and restlessness. Family members reported that another child from their household had died hours before, and within hours, the woman and her child also died. A team from IEDCR went to investigate on November 5.

All patients presenting to Gowainghat Health Complex with vomiting were referred to Osmani Medical College Hospital in Sylhet and an expert medical committee comprised of 11 experts from around the country was formed to evaluate patients. Seventeen cases, defined as any patient who had a history of vomiting followed by restlessness or unconscious state after November 2, were identified, with 8 deaths. Initial laboratory tests indicated that patients experienced severe liver damage; liver enzymes (alanine aminotransferase [ALT]) were high in 7 out of 11 patients tested (2,750 to 5,000 U/L). The team learned that many cases had eaten ghagra shak, an uncultivated plant, in the hours before falling ill. Blood, urine and vomitus samples were collected from patients and ghagra shak specimens and water samples from tubewells and ponds were collected from the affected village.

As the team completed a preliminary investigation and traveled back to Dhaka on November 8, 8 more children were admitted to Osmani Medical College Hospital with unconsciousness and a history of vomiting. The team immediately planned for more extensive investigations and on the following day, November 9, a combined IEDCR/ICDDR,B team went to Sylhet with two objectives: 1) to define the clinical syndrome of cases involved in the outbreak, and 2) to generate hypotheses about the etiology of disease and risk factors associated with illness.

Clinical syndrome

Eighty-one cases were identified from 11 different villages in Gowainghat and Companyganj upazilas through case finding efforts at hospital and in the community. Illness histories were collected for all cases and blood, urine, throat swabs, and when possible, cerebrospinal fluid was collected from living cases. Blood tests and MRIs were conducted when possible.

All cases occurred over a 10 day period from November 2-11 (Figure 1). The mean age was 17 and 68% (55/81) were female (Table 1).

All patients experienced vomiting, per the case definition, and of those 53% (43/81) also experienced altered mental status within hours of onset of symptoms, including disorientation and unconsciousness (35%). Twenty-four percent (19/81) of cases died, all of whom experienced altered mental status (Table 2). Eighty-four percent (16/19) of deaths were in children <15 years old. Local laboratory investigations performed on a subset of cases suggested that patients experienced significant liver damage.

Seventy percent (14/20) had elevated ALT levels (15 to 6,795 U/L) and 53% (9/17) had abnormal prothrombin times. Diagnostic tests for malaria, Japanese encephalitis, Nipah virus, and influenza were all negative; toxic poisoning was suspected as the cause of illness.

Xanthium strumarium with fruits (Source: Rain Forest Research Institute, Jorhat, Assam)

Anthropological investigation

The anthropological team spent one week in outbreak villages and collected illness history timelines, travel and food histories, and probed about exposures to sick animals and humans. Given the hypothesis that the outbreak was caused by toxic poisoning, the team questioned cases and their families about possible man-made chemical exposures, and recorded all goods which had been purchased or brought into affected households prior to the outbreak. They also focused on consumption of uncultivated plants based on evidence from India where outbreaks with similar clinical characteristics were caused by consuming toxic wild plants (1-3) and the initial reports that cases in this outbreak had consumed ghagra shak before they became ill.

Outbreak villages were located in remote areas in northern Sylhet District on the Indian border and case households were poor; many reported eating less than three meals per day. The household wage earners were day labourers and engaged primarily in collecting stones, fishing, and digging. Villagers reported that cases became ill suddenly with frequent and profuse vomiting, followed quickly (within minutes or hours) by unconsciousness for some. In the index household, the mother and her three children became ill within 6 hours of each other and all but one child died within 12 hours of onset of illness.

In total, the qualitative team collected in-depth food histories for 33 cases from 4 different villages, including 14 who died. Of those 33, 31 reported that they ate ghagra shak within 24 hours before they became ill. The team identified others in these villages who consumed ghagra shak at this time and experienced stomach upset or diarrhea, although they didn’t have vomiting or severe illness.

Locals described their usual ghagra shak collection and preparation practices for the team. Usually, they don’t eat ghagra shak seedlings as they consider them to be poisonous. Instead they eat the stems and stalks of older plans, which are believed to be safe, and discard the root and leaves. However, the ghagra shak plants consumed before the outbreak were smaller than usual for this time of year, due to late flooding. They reported that they relied upon uncultivated plants more these days, as severe and prolonged flooding earlier in the year prevented them from harvesting crops.

Exposure study

All household members in two outbreak villages were questioned about their food intake during the outbreak period and symptoms experienced to investigate exposures associated with illness. Appropriate proxy respondents were identified for children and persons hospitalized or deceased. Illness was defined as vomiting and unconsciousness the two days before the first death in the village through the day the last death occurred in the village. This definition was used because deaths and unconsciousness were memorable to villagers and onset of illness to death in cases usually occurred within hours. We estimated the association between exposures and illness by calculating relative risk ratios with 95% confidence intervals and considered results with a p<0.05 to be statistically significant.

Generalized estimating equations were used to investigate associations, by calculating prevalence ratios, between illness and exposures with significant relative risk ratios in order to account for the clustering of illness observed in households. One hundred thirty-one households were enrolled; 647 persons were interviewed (85% of all residents). Individuals not interviewed were unavailable to be contacted on multiple visits to the household. Twenty-six persons experienced vomiting and unconsciousness during the outbreak period. In univariate analysis, 7 exposures were associated with illness and persons who reported eating ghagra shak during the outbreak period were 14.7 times (risk ratio 95% confidence interval, 7.4-29.5, p<0.001) more likely to experience vomiting and unconsciousness than those who did not eat the plant. When household clustering was controlled for using generalized estimating equations, only two exposures were associated with illness. The odds of developing illness after eating gaghra shak (28.9, 95% CI 9.2-90.8, p<0.001) and keshari lentils (17.5, 95% CI 3.1-99.8, p<0.001) remained strong; however, only 19% (5/26) of persons who developed illness reported eating the lentils compared to 46% (12/26) who reported consuming ghagra shak.

Reported by: Infectious Disease and Vaccine Sciences Programme, ICDDR,B andInstitute for Epidemiology, Disease Control and Research, Ministry of Health and Family Welfare, Government of Bangladesh Supported by: Centers for Disease Control and Prevention, USA, and Government of Bangladesh

Comment

Evidence suggests that this lethal outbreak (case fatality ratio 24%) was caused by toxic poisoning from ghagra shak, a local uncultivated plant. Ghagra shak is the local plant name for Xanthium strumarium (4) and consumption of Xanthium species leaves or seeds have been associated with illness and death in livestock and an outbreak of toxic poisoning in children in Turkey (5-8). Carboxyatractyloside has been found in immature plants and seeds and is believed to be the toxic agent responsible for causing illness (8). Eighty four percent of deaths in this outbreak were in children and 64% of cases were women; they were likely at increased risk of poisoning because of their relatively lower body weight.

The clinical syndrome of cases is consistent with toxic poisoning; onset of symptoms to death was rapid and liver function tests were distinctly abnormal in many patients. Other tests were normal, including an MRI done on one patient with altered mental status, and suggest that these symptoms were not caused by an infectious agent. Cases were identified from many different remote villages in northern Sylhet; persons in other remote villages may have also fallen ill but not sought care, and thus some cases may remain unknown. The epidemiologic data show a strong association between consuming ghagra shak and developing vomiting and altered mental status during this outbreak.

This outbreak is further example of how poverty puts individuals at higher risk for disease and death. Villagers reported that they deviated from their usual ghagra shak consumption practices this year because their crops were destroyed by severe and late flooding and they were unable to purchase enough food from the markets. Flooding also contributed to the immaturity and toxicity of ghagra shak plants available; studies have shown that the leaves are only toxic during the cotyledon stage (9). Public health messages against eating ghagra shak should be developed and widely disseminated to prevent a recurrence of similar outbreaks next fall. Follow-up studies to assess the true magnitude of the outbreak and the role of uncultivated plants in the diets of rural Bangladeshis should be considered. Eating uncultivated plants has been associated with outbreaks of similar disease in India during times of food scarcity (1-3). Physicians should be alerted to clusters of patients presenting with vomiting and altered mental status, especially during years and seasons when access to food is limited. Any suspect clusters of illness should be immediately reported to health officials.

(Published with permission from “Health and Science Bulletin”, Volume 6,  Number 2, June 2008; ISSN 1729-343X)

References:

1. Vashishtha VM, Kumar A, John TJ, Nayak NC. Cassia occidentalis poisoning as the probable cause of hepatomyoencephalopathy in children in western Uttar Pradesh. Indian J Med Res 2007;125:756-62.

2. Vashishtha VM, Kumar A, John TJ, Nayak NC. Cassia occidentalis poisoning causes fatal coma in children in western Uttar Pradesh. Indian Pediatr 2007;44:522-5.

3. Vashishtha VM, Nayak NC, John TJ, Kumar A. Recurrent annual outbreaks of a hepato-myo-encephalopathy syndrome in children in western Uttar Pradesh, India. Indian J Med Res 2007;125:523-33.

4. Chowdhury MSU. Bangladesh: Coutry report to the FAO international technical conference on plant genetic resources. Leipzig: Food and Agriculture Organization, 1995. 112 p.

5. Mendez MC, dos Santos RC, Riet-Correa F. Intoxication by Xanthium cavanillesii in cattle and sheep in southern Brazil. Vet Hum Toxicol 1998;40:144-7.

6. Stuart BP, Cole RJ, Gosser HS. Cocklebur (Xanthium strumarium, L. var. strumarium) intoxication in swine: review and redefinition of the toxic principle. Vet Pathol 1981;18:368-83.

7. Turgut M, Alhan CC, Gurgoze M, Kurt A, Dogan Y, Tekatli M, et al. Carboxyatractyloside poisoning in humans. Ann Trop Paediatr 2005;25:125-34.

8. Witte ST, Osweiler GD, Stahr HM, Mobley G. Cocklebur toxicosis in cattle associated with the consumption of mature Xanthium strumarium. J Vet Diagn Invest 1990;2:263-7.

9. Cole RJ, Stuart BP, Lansden JA, Cox RH. Isolation and redefinition of the toxic agent from cocklebur (Xanthium strumarium). J Agric Food Chem 1980;28:1330-2.