While self-poisoning with Pesticides is the most common method of self-harm in Asia, self-poisoning with plant seeds or fruits is also common, especially in the South Asian region [1]. While most deaths follow ingestion of Thevitia peruviana ( yellow oleander) seeds, other locally common plants are also implicated [2]. We noted that cases of self-poisoning with Cerbera manghas (sea mango, pink eyed cerbera, odollam tree) fruits, while being most commonly reported in Tamil Nadu and Kerala [3,4] are seen in the Eastern Province of Sri Lanka.

Cerbera manghas is a poisonous plant belonging to the Apocynaceae family notorious for its cardiac glycoside cardiotoxicity. C. manghasis called diyakaduruin Sinhala, kattuarali in Indian Tamil; natchuchukkaiin Sri Lankan Tamil. Cerbera manghas Linn (previously also Cerberaodollam Gaertn) is common along the coasts of south Asia and south East Asia, Northern Australia, and Polynesia. In Sri Lanka, especially along coastal regions, sea mango (Cerbera manghas) is a very common plant and readily accessible. It grows to a height of 5 to 10 metres and bears whorled branchlets with terminal lanceolate leaves. The fruit of the sea mango turns bright red at maturity and appear very much like the edible mango (Mangifera spp.) fruit. The latex present throughout the plant contains cardenolides (such as cerberin, neriifolin, and cerberoside) (Figure 1) that cause vomiting, cardiac dysrhythmias, and hyperkalemia via inhibition of the Na⁺/K⁺ATPase.

Figure 1: Cerbera odollam tree leaves and seeds.

Although widespread in Sri Lanka [5], and well known to be poisonous [6], only a one report has been published from the island. However, there have been no publications reporting clinical data from poisoned patients.

The aim of this study was to evaluate the clinical manifestations, biochemical findings and outcome of management using currently available treatment in the poisoning unit of a tertiary care hospital in Eastern Sri Lanka.

Study population

Patients with Cerbera manghas poisoning admitted to tertiary care hospital in Eastern Sri Lanka from 1st January 2011 to 31st October, 2013 by were included using predesigned questionnaires, retrospectively.

Treatment of patients

Patients with Cerbera manghas poisoning were treated in medical wards and poisoning unit, tertiary care hospital in Eastern Sri Lanka with different treatment modalities such as activated charcoal, intravenous boluses of atropine/isoprenaline, temporary pacing and supportive measures depending on the clinical manifestations.

Electrocardiographic monitoring

12-lead standard electrocardiography (INNOMED Medical ECG machine) and 2-lead ECG monitoring were taken during the standard work up of each patient in this unit.

Blood samples

Five milliliters (5 ml) venous blood was collected from each patient. Serum was separated by centrifuging in a laboratory centrifuge at for three minutes after blood clotting and retraction at room temperature. Serum potassium (K⁺) and Sodium (Na⁺) were analyzed at the Department of Chemical Laboratory. Renal function including blood urea nitrogen [BUN], creatinine and liver function indices including SGOT, SGPT, PT and serum protein were analyzed in the clinical laboratory of Batticaloa teaching hospital using standard automated techniques.

Statistical analysis

Differences between the two groups (patients with significant arrhythmia vs. patients with insignificant arrhythmia) were analyzed with pair-wise comparisons. Baseline results are presented as counts and percentages and as mean ± SD for continuous variables. A P value <0.05 was considered significant.

We noted 583 cases of poisoning from 1st January, 2011 to 31st October, 2013. Among these, 192 (32.93%) cases were due to Yellow Oleander poisoning, while 48 (8.23%) cases were due to Cerbera manghas poisoning. There were 48 patients [Mean age: 21 (± 0.43) yrs], (Male: Female=35:13) with Cerbera manghas self-poisoning. The great majority were in the 16-30 year age group (40, 83.3%). The ethnicity of the majority of patients was Tamil (42, 87.5) (Table 1). Twenty four (50%) of them had ingested half a seed (one seed weighted about 20 g) (Table 2).

Table 1: Personal characteristics of patients.

Table 2: Number of seeds ingested by patients.

Most of the patients were symptomatic with symptoms of cardiac glycoside toxicity including vomiting (48, 100%), dizziness (24, 50%) and abdominal pain (20, 41.6%). Abdominal pain and diarrhea alone was rare presentations. Neurological manifestations such as drowsiness and restless (8, 16.6%) were also associated. Cardiac dysrhythmias such as bradycardia or an irregular pulse were the examination findings. ECG findings showed 1st, 2nd, 3rd degree heart block and sinus bradycardia (Table 3). A normal ECG was found only in 6 patients. Atrial fibrillation was the tachyarrythmias. Mean serum potassium concentration was significantly higher in patients with significant cardiac arrhythmias that required specific management transfer to CCU and temporary pacemaker insertion.

Table 3: ECG findings of Patients.

Fourteen of them had higher serum potassium concentrations (6.4 (± 0.97) mmol/L). Ten patients had life threatening hyperkalemia [7.4 (± 0.17)]. All patients developed cardiac toxicity and hyperkalemia within 24 hrs of ingestion of seeds. These severe cardiac toxicity and hyperkalemia were observed among patients (16, 33.3%) having half seed of ingestion. All other blood investigations including renal and liver markers are within normal range (Table 4).

Table 4: Blood Investigations of patients.

All patients were treated with multiple doses of activated charcoal. Patients with bradyarrhythmias were treated with intravenous atropine (18, 37.5%) and intravenous infusions of isoprenaline (4, 12.5%). Temporary cardiac pacing was required in nine of them who had not responded to drug therapy. There were eight deaths who had both 3rd heart block and high serum potassium concentrations (>7.0), (Figure 2; ECG tracings). Three died before definitive treatment could be instituted. The majority of patients with cardiac toxicity had an uneventful recovery. Multiple doses of activated charcoal alone were safe and adequate in most cases in our study. However it was not statistically proven. All patients with hyperkalemia were managed only with insulin-dextrose (10 units Insulin with 50 ml 50% Dextrose regime).

Figure 2: ECG tracing of a patient; 54 year old man presented with severe vomiting, abdominal pain and restless following ingestion of half seed of cerbers manghas after 8hrs of ingestion. His ECG showed complete heart block. His serum Potassium was 7.9 mmol/dL. He died within 2 hrs of hospital admission even before pacing was instituted.

Features of severe toxicity such as persistent vomiting, severe abdominal pain, neurological signs and persistent hyperkalemia were significantly associated with high risk of mortality and morbidity (P<0.05). The risk of cardiac toxicity was observed with patients with ingestion of less amount of seeds (half seed).

A previous study in 2000/2001 showed that Cerebera manghas is the cause for 20% deaths from plant poisoning and 4.4% of all self-harm deaths in this hospital on the eastern coast of Sri Lanka [7]. It was the second commonest plant poisoning in Eastern Sri Lanka next to Yellow Oleander Poisoning in our study period. Review of our study confirmed that Cerbera manghas plant poisoning produces clinical features typical of cardenolide poisoning [8]: in particularly; vomiting, cardiac dysrrhythmias and hyperkalemia. Electrocardiographic changes are the most obvious marker of Cerbera manghas toxicity [9]. The most common ECG abnormality is bradycardia, which may be sinus bradycardia, sinus arrest or exit block, or AV nodal dissociation (third degree heart block).

Management involved gastric lavage, multiple doses of activated charcoal, administration of atropine to counter bradycardia and Insulin-Dextrose to treat hyperkalemia. The effect of atropine on outcomes in cases of bradyarrhythmia has not been formally assessed. However a study from Trivandrum reported a poor outcome if bradycardia did not respond to atropine and atropine was not effective [10].

Cardic arrhythmias and persistent hyperkalemia are important predictable mortality risk for Cerbera manghas plant poisoning. There were eight deaths (16.67%) due to third-degree heart block and life threatening hyperkalemia in our study. These deaths were observed among patients having half seed of ingestion. These severe cardiac toxicity and hyperkalemia were observed among patients having half seed of ingestion and not vomiting seeds immediately. This may be due to rapid absorption of seeds in gastrointestinal tract following ingestion. The patients taking more seeds vomited immediately within 2 hrs. As a result, they have less cardiac toxicity compared those having lees amount of ingestion of seeds.

Despite its relative uncommon prevalence compared to Yellow Oleander poisoning, it is much feared by the medical staff due to high case fatality and the lack of pacemakers or anti-digoxin Fab for therapy [11].

A limitation of our study is also same as in previous study [7] that neither blood nor gastric contents were tested for Cerbera manghas cardenolides to prove ingestion. The responsible fruit ingestion was identified only from history from patient and relatives. However, Cerbera manghas is very well known in the region and easily recognized, increasingly the like hood that the ingested fruit would be accurately identified.

Cerbera manghas self-poisoning was common among young males in Batticaloa district of the Eastern Province. The cardiac toxicity observed within 24 hrs of ingestion of seeds. The severe toxicity was associated with less amount of seeds. Most patients had specific symptoms of Cerbera manghas poisoning. AV conduction defects are common. Patients presenting with features of severe toxicity such as persistent vomiting, severe abdominal pain, neurological signs and persistent hyperkalemia has higher risk of mortality.

Cerbera manghas poisoning is the second commonest plant poisoning in Eastern Sri Lanka but lack of resources can significantly impair the successful management in the hospital setting. We would like to emphasize the importance of equipping these hospitals with adequate resources, e.g: pacemakers for therapy Cerbera manghas poisoning to prevent adverse outcomes.

We thank the staffs of Department of Emergency, Cardiology and Poisoning Units, Teaching Hospital, Batticaloa for the help during the course of these studies. We also thank Prof. M. Eddleston for his data from his previous paper published in 2008.