Acute hyperglycemic emergencies in children with type 2 diabetes

Not long ago, diabetic ketoacidosis (DKA) was thought to be pathognomonic of type 1 diabetes in children. However, we now know that children with type 2 diabetes may also present with DKA (1,2). More recently, hyperosmolar hyperglycemic syndrome (HHS) has also been reported in children with type 2 diabetes (3). These two life-threatening hyperglycemic emergencies are well known conditions now occurring in a new group of children. Correct diagnosis is critical because the principles of treatment may be different in these conditions, especially in terms of the rate of fluid administration. DKA is defined as a pH of less than 7.3 and serum bicarbonate under 15 mmol/L, with random blood glucose over 11.1 mmol/L. It is often impossible to confirm this diagnosis in rural and remote communities because of the lack of availability of capillary blood gases. ‘Large’ urine ketone levels indicate that there is ketonemia but do not indicate whether there is a systemic acidosis. Alternative surrogate diagnostic measures are total venous carbon dioxide or capillary beta-hydroxybutyrate level using a portable blood glucose/ketone meter to indicate the severity of the ketonemia. DKA in type 2 diabetes is due to relative insulin deficiency in a person with longstanding hyperglycemia, and is often exacerbated by a sudden increased demand for insulin during an intercurrent illness. There may have been an earlier prolonged period of asymptomatic insulin resistance with eventual hyperglycemia due to pancreatic beta cell failure and insulin deficiency. This explains why insulin levels are not helpful in the classification of diabetes type; a low serum insulin level does not discriminate between type 1 and type 2 diabetes (4). Clinical features of insulin resistance remain the most important diagnostic tools in differentiating type 1 and type 2 diabetes. The most important of these diagnostic tools are a strong family history of type 2 diabetes, especially if the mother had prepregnancy or gestational diabetes, or if the patient is from a high-risk ethnic background or had childhood obesity with acanthosis nigricans. First Nations children in central Canada, especially those with Oji-Cree heritage, appear to have the highest risk of type 2 diabetes (5). In some cases, negative islet-specific antibody levels are needed to confirm type 2 diabetes (4). Children with type 2 diabetes and DKA will not always have classical symptoms of hyperglycemia preceding the acute illness. They may only complain of increasing abdominal symptoms due to ketonemia. The current recommended management of DKA in children with type 2 diabetes is the same as in type 1 diabetes: careful, judicious replacement of fluid, insulin and potassium to avoid cerebral edema, hypokalemia and hypoglycemia (6). Fluid infusion rates with normal saline are slower in DKA in children than adults, with rates calculated to replace the fluid deficit over 48 h to avoid cerebral edema. Children with type 1 diabetes have a higher risk of cerebral edema than adults when recovering from DKA (7). The diagnostic triad of HHS is severe hyperglycemia (plasma glucose over 33 mmol/L), hyperosmolarity (serum osmolarity over 320 mOsm/kg) and mild ketoacidosis (bicarbonate over 15 mmol/L) (Table 1). The classical picture of HHS is an elderly debilitated person with type 2 diabetes who develops severe dehydration rather than acidosis. This may also be seen in an infant with new-onset type 1 diabetes who has limited access to water. Severe dehydration is the cardinal feature of HHS. In adolescents, longstanding unrecognized hyperglycemia with high intake of sugar-containing drinks may contribute to the slow development of HHS. At presentation, most subjects are severely dehydrated and often have total body water loss of 15% to 20%, or 12% of body weight, compared with total body water loss of 10% in DKA. In adults with HHS, rapid correction of intravascular volume and shock associated with HHS is recommended (8). It seems prudent to consider children with HHS to have a similar risk of cerebral edema as children with DKA. However, it must be noted that severe hyperosmolarity may mask a severe degree of dehydration and increase the risk of multiorgan failure. In selected cases, aggressive fluid management may be necessary and should only be performed in a tertiary paediatric intensive care setting. The risk of death in children with HHS is unknown, and children rarely have serious comorbid conditions, as is the case in elderly adults. TABLE 1 Diagnosis of diabetic ketoacidosis (DKA) versus hyperosmolar hyperglycemic syndrome (HHS) in children A mixed picture of DKA and HHS may occur with severe hyperglycemia and hyperosmolarity associated with ketoacidosis in type 2 diabetes. There have been three published series in children. In a series of seven children, all of the children were initially thought to have type 1 diabetes and DKA, but all died of HHS and type 2 diabetes (9). In another series of four children, all four had a mixed picture of HHS and DKA, and two of the four children died (3). In the authors’ series of 13 First Nations children with type 2 diabetes and DKA (1), none died despite the fact that seven of 13 children presented with severe DKA (pH of 7.0 or less). On closer inspection of the biochemical results, we now recognize that five of 13 children had a random plasma glucose level of over 33.3 mmol/L, indicating a mixed picture with HHS and DKA. At this time, the optimum treatment of HHS in children with type 2 diabetes to reduce the risk of shock and death is unknown. The optimum treatment of DKA in children with type 2 diabetes to reduce the risk of cerebral edema is also unknown. Perhaps it is time to plan a careful, multicentre study of a treatment protocol in children with life-threatening presentations of type 2 diabetes to identify the optimum fluid management for each condition. In the meantime, it is prudent to follow conservative recommendations for the treatment of DKA in children (6). In transferring these children from remote communities to paediatric centres for intensive medical care, intravenous administration of normal saline at no more than twice the normal fluid maintenance rate is vital; intravenous (or subcutaneous) insulin and potassium are not necessary in transit because there is no immediacy for either therapy, and both are potentially dangerous. The most effective strategy for preventing the acute complications of DKA or HHS in children with type 2 diabetes is an effective screening program for type 2 diabetes in children at risk. The 2003 Canadian Diabetes Association clinical practice guidelines (10), endorsed by the Canadian Paediatric Society’s First Nations and Inuit Health Committee (11), recommend a fasting plasma glucose every two years in obese children 10 years of age or older if they have two additional risk factors. Ethnicity and a parent with type 2 diabetes are two of these risk factors. Earlier and more frequent screening should be considered in obese Aboriginal children with acanthosis nigricans and a very strong family history of type 2 diabetes, particularly if the mother had prepregnancy type 2 diabetes. In summary, HHS and DKA occur in children with type 2 diabetes and are potentially deadly. Little is known about the etiology, optimum treatment and outcomes in this age group, but the current consensus is that they should be managed in a carefully monitored setting with a conservative fluid regimen, as in DKA children with type 1 diabetes. All children should be referred to a centre with expertise in managing these life-threatening conditions; paediatric centres, in turn, have a responsibility to participate in multicentre trials to provide new evidence to increase our knowledge. The 2006 Canadian Paediatric Surveillance Program surveillance project on non-type 1 diabetes will give a more accurate picture of incidence rates of type 2 diabetes, and possibly its presentation with HHS or DKA in children across Canada (12).