Risk Factors for Adverse Neurodevelopment in Transient or Persistent Congenital Hyperinsulinism

Abstract

Objective: Aim was to identify hypotheses why adverse neurodevelopment still occurs in children with transient or persistent hyperinsulinism despite improvements in long-term treatment options during the last decades.

Material and methods: A retrospective review of 87 children with transient (n=37) or persistent congenital hyperinsulinism (CHI) (n=50) was conducted at the University Children's Hospital Duesseldorf, Germany. Possible risk factors for neurodevelopmental sequelae due to hypoglycemia were analyzed with a focus on the first days after onset of disease.

Results: Median age at follow-up was 7 years (IQR 8). Adverse neurodevelopmental outcome was seen in 34.5% (n=30) of all CHI patients. Fifteen had mildly abnormal neurodevelopment and 15 had a severe hypoglycemic brain injury. In univariate analysis, mildly abnormal neurodevelopment was associated with the diagnosis of persistent CHI (odds ratio (OR) 8.3; p=0.004) and higher birth weight (mean difference 1049 g; p<0.001). Severe hypoglycemic brain injury was associated with the diagnosis of persistent CHI (OR 5.1; p=0.013), being born abroad (OR 18.3; p<0.001) or in a lower-level maternity hospital (OR 4.8; p=0.039), and of note history of hypoglycemic seizures (OR 13.0; p=<0.001), and a delay between first symptoms of hypoglycemia and first blood glucose measurement/initiation of treatment (OR 10.7; p<0.001). Children with severe hypoglycemic brain injury had lower recorded blood glucose (mean difference -8.34 mg/dl; p=0.022) and higher birth weight than children with normal development (mean difference 829 g; p=0.012). In multivariate binary logistic regression models, lowest blood glucose <20 mg/dl (OR 134.3; p=0.004), a delay between initial symptoms and first blood glucose measurement/initiation of treatment (OR 71.7; p=0.017) and hypoglycemic seizures (OR 12.9; p=0.008) were positively correlated with severe brain injury. Analysis showed that the odds for brain injury decreased by 15% (OR 0.85; p=0.035) if the blood glucose increased by one unit.

Conclusion: While some risk factors for adverse outcome in CHI are not influenceable, others like lowest recorded blood glucose values <20 mg/dl, hypoglycemic seizures, and insufficiently-or even untreated hypoglycemia can be avoided. Future guidelines for management of neonatal hypoglycemia should address this by ensuring early identification and immediate treatment with appropriate escalation steps.

Keywords: brain injury; hyperinsulinism; hypoglycemia; neurodevelopment; outcome; risk factors.

Figure 1

Birth weight according to neurodevelopmental outcome. Values are displayed as mean ± standard deviation. ANOVA with Tukey post hoc comparison showed significant differences in the mean for normal neurodevelopmental outcome (n = 50) vs. mildly abnormal development (n = 14) (2,892 ± 779 vs. 3,940 ± 1,013, p = <0.001) and normal neurodevelopmental outcome vs. severe brain injury (n = 13) (2,892 ± 779 vs. 3,720 ± 1,220, p = 0.012). *p = < 0.05. n.s., not significant.

Figure 2

Lowest blood glucose values according to neurodevelopmental outcome. Values are displayed as mean ± standard deviation. ANOVA with Tukey post hoc comparison showed significant differences in the mean for normal neurodevelopmental outcome (n = 53) vs. severe brain injury (n = 14) (20.6 ± 11.0 vs. 12.3 ± 8.6, p = 0.022). Comparison between the remaining groups showed no significant differences. *p < 0.05, n.s, not significant.