The use of exenatide in severely burned pediatric patients

Hyperglycemia is a common finding in critically ill patients that has been associated with increased morbidity and mortality. In burns, it has also been shown that hyperglycemia is deleterious. Gore et al. [1] found that hyperglycemia was associated with an increased rate of muscle protein catabolism [2] and increased morbidity and mortality in this population. In addition, Hemmila et al. [3] found that IIT was associated with a lower incidence of pneumonia, ventilator-associated pneumonia and urinary tract infections; and Pham et al. [4] reported similar findings and a positive association of IIT with survival rates following IIT treatment in pediatric burned patients. Our group has recently shown that IIT in severely burned pediatric patients was associated with improved post-burn morbidity, such as infection, sepsis, and organ function [5].

However, the Normoglycemia in Intensive Care Evaluation and Survival Using Glucose Algorithm Regulation (NICE-SUGAR) found no benefit and an increased incidence of hypoglycemia with IIT in adult critical care [6, 7]. However, detailed analysis of the NICE-SUGAR study suggests a better outcome within the trauma subpopulation with intensive insulin treatment [6].

Since current evidence supports the use of IIT in trauma patients, the study of novel therapies to decrease hyperglycemia in burn patients without increasing the risk of hypoglycemia is warranted [8]. Incretin-based therapies are currently among the newest classes of available glucose-lowering agents [9]. The incretin effect consists of higher insulin production after an oral ingestion of glucose than after an intra-venous load one [10]. The incretins that have been identified are glucose-dependent insulinotropic peptide (GIP) and glucagon-like peptide-1 (GLP-1). Exogenous GLP-1 has been shown to reduce glucose concentration when administered to hospitalized patients [11, 12]. Activation of the incretin receptors on β-cells increases insulin release in response to glucose [13] and may have additional beneficial effects, as it has been suggested that these drugs promote enhanced glucose disposal in peripheral tissues and protect against ischemia/reperfusion injury [14].

Exenatide is a synthetic peptide originally identified in the lizard Heloderma suspectum [15] that possesses incretin-mimetic actions including suppression of glucagon secretion and delay of gastric emptying [16]. It has been shown to bind and activate the human glucagon-like peptide-1 (GLP-1) receptor in vitro [17]. We hypothesized that exenatide would decrease the amount of exogenous insulin administered to the patients and the incidence of hypoglycemia in the acute setting of severely burned pediatric patients.

Six patients were randomized to the treatment group and 18 patients to the IIT group (Figure 1). Demographics showed no significant difference between both groups (Table 1). Glucose values were similar in both groups (Table 2). The number of glucose determinations was similar between both groups: 143 ± 38 for the exenatide group and 139 ± 59 for the IIT group (P = 0.79).

Longitudinal analysis of all the glucose values (6 AM, daily, maximum and minimum glucose levels) along the acute stay showed similar values in both groups (Figure 2). The incidence rate of days with less than three glucose determinations per acute stay was 15.5 days/30 days for the IIT group and 14.5 days/30 days in the exenatide group (P > 0.05).

Patients in the IIT group received significantly more insulin to maintain similar glucose levels when compared to the exenatide patients (P = 0.01), while serum insulin levels (endogenous and exogenous) were not significantly different between both groups (P > 0.05) (Figure 3). Three patients in the exenatide group did not receive exogenous insulin, while all the patients in the IIT group received insulin.

The MAGE was not statistically different between both groups (P = 0.61) (Figure 4). Similarly, % CV was similar in both groups (Figure 5).

The incidence rate of hypoglycemia was similar in both groups (0.38 events/patient-month). There were 17 events of moderate hypoglycemia (40 to 59 mg/dl) and 1 of severe hypoglycemia (< 40 mg/dl) in the IIT group and 6 events of moderate hypoglycemia and none of severe hypoglycemia in the exenatide group.

The REE [4] was not significantly different between the groups (Figure 6). The total amount of calories administered to the patients is shown in Figure 7. It was not different between groups.

Exenatide was well tolerated by all the patients and no adverse reaction associated with the administration of exenatide was documented. There was no mortality in any of the groups.

Incretin-based drugs are a family of substances that lower glucose levels by enhancing glucose-dependent secretion of insulin, a condition known as the incretin effect [24]. Incretin-based drugs require a hyperglycemic state for their effects [25], thus, hypoglycemia is uncommon even with high-doses of the drugs [26]. This and many other described effects of GLP-1 analogues, such as increased endogenous insulin secretion and peripheral glucose uptake and protection against ischemia/reperfusion, made these drugs particularly suitable for burn patients. Other drugs may decrease glucose levels with a low risk of hypoglycemia or may increase peripheral glucose uptake, but in theory with GLP-1 analogues we would achieve these effects with only one drug.

Among the incretin-based drugs, we decided to administer exenatide because it is known that it binds with equal affinity to the GLP-1 receptor and produces similar glucose-lowering actions to GLP-1 [27]. Additionally, exenatide is known to increase insulin-dependent glucose uptake in muscle and fat while GLP-1 does not have such action [28]. Furthermore, its longer half-life allows for subcutaneous administration q- 12 h facilitating the management of severely burned patients in the ICU. According to its described effects, we expected that patients receiving exenatide treatment would receive less exogenous insulin. Since hyperglycemia was treated with administration of exogenous insulin in both groups to achieve normal levels, we did not expect significantly different glucose levels between patients, as observed.

Variability of glucose has been associated with increased mortality and has been considered one of the most important goals of glucose management in ICU [29]. We, therefore, included analysis of MAGE. Although mean MAGE was lower in the patients receiving exenatide, with the actual sample size it is not possible to make any conclusion on this regard. We found no difference in the incidence of hypoglycemic events in both groups, probably as a result of the insulin administration with a target glucose level of 80 to 140 mg/dl in all patients. By modifying the target glucose level, we hypothesize that the incidence of hypoglycemia would decrease in these patients. Further studies are necessary to determine the safest target glucose that improves morbidity and mortality in the burn population.

This study was designed as an open-label pilot study and was one of the multiple trials we conduct at our institute. At our institute, we have several clinical trials to study the effect of anti-catabolic, anabolic, and glucose modulation agents. The primary aim of the present study was to assess efficacy and feasibility (that is, effects on insulin and glucose metabolism) in a limited patient population. Based on previous efficacy studies at our institute, we hypothesized that we would need about six patients in the intervention group. In order to decrease variability and decrease the low sample error, we randomized three times as many patients (18 patients) to the control group. We are clearly underpowered to draw any conclusions on the effect of GLP-1 on morbidity or mortality. However, we demonstrated in the present open label trial that GLP-1 appears to be a safe adjunct therapy to achieve glucose in a pediatric burn population.

Exenatide was originally described as a therapy for adult diabetic patients, and although its use in pediatric patients has been suggested [30, 31], the available literature about its use and safety in these patients is scarce at best. It has been described that exenatide may cause some adverse effects that can be severe (for example, pancreatitis) [32]. However, we did not observe complications or any undesirable effect associated with exenatide administration in the studied patients (for example, gastrointestinal symptoms, injection site reactions and hypersensitivity).

This study has limitations due to the small number of patients and the fact that the treatment was not blinded for safety reasons. A larger sample size would be necessary to assess other aspects, such as true complication rate, variability of glucose and the effect of exenatide on hypoglycemia and REE. The percent predicted REE was lower (closer to normal) in the first two weeks post admission in the exenatide patients, although not statistically significant. The absorption of exenatide after subcutaneous administration might be inconsistent in edematous patients. It has been shown that oral glutamine increases circulating GLP-1 [33] potentially stimulating endogenous GLP-1 secretion in the IIT group. This might have affected the glucose metabolism in the IIT group. Plasma concentration of exenatide and GLP-1 should be included in future studies. It would have also been interesting to determine C-peptide in plasma to assess endogenous insulin production and glucagon concentrations and we recommend doing so for future studies. However, this study was the first attempt to find alternate glucose lowering agents in the setting of burn critical care. We believe that the utility of alternative drugs, (for example, peroxisome proliferator-activated receptors (PPAR) agonists), needs to be assessed since many of these agents are less expensive and as equally safe as GLP-1. Therefore, a study assessing the utility of these drugs in the burn population is warranted. In this trial, we found that exenatide was effective, safe, and well tolerated by the severely burned pediatric population.

Administration of exenatide to severely burned pediatric patients reduced the amount of administered exogenous insulin and was well tolerated during their acute setting. The GLP-1 analogue tested in this trial appears to be safe and reliably modulates glucose in these patients.