We present a case of severe AN with sudden death early in the refeeding phase, with findings of cerebellar tonsillar herniation and strong cerebral edema on autopsy. To the best of our knowledge, this is the first report of the development of cerebellar tonsillar herniation associated with sudden death in a patient with AN. This report is significant because it highlights the severity of the markedly malnourished patient, such as those with severe AN, and has implications for the future monitoring of patients during refeeding. Here, we discuss the cause of death in our case and the cause of the cerebral edema that led to cerebellar tonsillar herniation.
Based on the clinical course and autopsy findings, the direct cause of death in this case may have been cardiogenic or neurogenic with cerebral edema resulting in tonsillar herniation. Both heart damage and cerebral edema could be the cause of death or the result of some conditions.
The patient's ECG at the time of admission suggested that she had suffered some cardiac damage. Turillazzi et al reported that “colliquative myocytolysis” is a possible histological hallmark of congestive heart failure in underweight AN patients [13]. In our case, there were no similar findings on cardiac histology. Although the patient had mild QTc prolongation on admission, we did not find further prolongation of the QTc on posthospitalization ECG monitoring, and it is questionable whether this directly led to lethal arrhythmia and sudden death.
It is known that when refeeding severely undernourished patients, electrolytes are transferred into the cells due to an increase in insulin concentration caused by elevated blood glucose levels, resulting in a decrease in serum phosphorus, potassium, and magnesium levels, and this condition is known as refeeding syndrome [2, 9–12]. Many published guidelines recommend an initial calorie level of 10 kcal/day for high-risk patients and 5-10 kcal/day for very high-risk patients to prevent RS [2, 9–12, 14]. Although the initial calories in the present case were within the recommended range, it is possible that the patient had a sudden change due to fatal arrhythmia.
On the other hand, the autopsy findings of this case showed unexpectedly massive general cerebral edema and tonsillar herniation. There have been no reports of patients with AN developing cerebral edema leading to tonsillar herniation. One autopsy case of AN with extensive cerebral edema following congestive heart failure was reported, but the direct cause of death in this reported case was attributed to pulmonary congestion associated with congestive heart failure [13]. Sundström et al reported a case in which the patient suffered from a disturbance of consciousness during the refeeding period from prolonged malnourishment due to causes other than AN, and intracranial pressure monitoring (ICP) showed a marked increase in intracranial pressure. In their case report, the initial calorie level was 1800 kcal/day, which was higher than recommended and suggested to be a risk factor for cerebral edema [14]. The initial calorie levels in the present case were within the optimal range and cannot be considered a similar factor.
In our case, the patient drank 700 ml of water in the 6 hours after admission. It is known that the regulatory systems related to sodium and water, such as ADH secretion and the renin-angiotensin-aldosterone system, are impaired in those with weight loss due to AN [15–17]. Similar abnormalities in sodium and water regulation have been reported in a rare case of cerebral edema associated with hyponatremia and subsequent cerebral herniation in a marathon runner, and water intoxication may have affected the development of cerebral edema in this case [18]. Although various guidelines for the prevention of refeeding syndrome do not provide definitive fluid restriction, further study may be needed in the future for extremely high-risk patients [2, 9–12].
The most common cause of sudden death in AN, with the exception of suicide, is fatal arrhythmia, and for this reason, during refeeding, attention has been given to monitoring biochemical tests and continuous ECG monitoring with an emphasis on arrhythmia [2, 3, 8, 19, 20]. However, few patients show evidence of arrhythmia, and fewer patients undergo autopsy. Therefore, it is not clear whether the arrhythmia is caused by electrolyte abnormalities or whether it is preceded by other fatal states, such as intracranial hypertension, respiratory arrest from cerebral herniation, and arrhythmia from hypoxia. In some cases, patients with severe AN may not respond to calls due to psychological resistance, and it is not easy to ascertain their state of consciousness, especially when they are asleep. Therefore, it is assumed that in many cases, as in this case, it may be difficult to determine whether the patient had a sudden disturbance of consciousness before the arrhythmia.
Accurate monitoring of intracranial pressure requires an invasive procedure and, given the rare frequency, should not be performed in malnourished patients without impaired consciousness [21]. However, given the risk of impaired consciousness and respiratory arrest even in the absence of arrhythmias, it may be worthwhile to study the necessity of repeated conscious status checks and continuous/nocturnal SpO2 monitoring in the acute phase of severely malnourished patients to prevent sudden death. In addition, in the unfortunate event of sudden death in patients with AN, autopsy and autopsy imaging may help to determine the frequency of massive cerebral edema in such situations.