Although preventable with earliest possible diagnosis and treatment, the frequency of neurodevelopmental impairment (NDI; brain damage) in congenital hyperinsulinism (HI) still reaches 30-50%, including delayed psychomotor development, cognitive impairment, cerebral palsy, epilepsy and blindness. The incidence of transient HI ("perinatal stress hyperinsulinism") has been estimated to 1:1,200, whereas persistent HI occurs in 1:10,000 to 1;40,000. Both transient and persistent hyperinsulinism, here together named pathological HI, impose, unfortunately, a high risk of NDI if not promptly diagnosed and treated.
According to the classical WHO consensus, an ideal universal screening should be performed with a suitable test to diagnose a condition early enough to institute an effective treatment. Moreover, the screening must be specific, affordable, harmless, and worldwide applicable.
From a yet unpublished retrospective pilot study, a novel proposal has been established of how to implement timely screening by use of umbilical cord blood glucose (UCBG), which is already routinely measured in cord gas analysis for asphyxia, to predict neonatal HI. Until today, the use of UCBG screening for neonatal hypoglycemia is virtually a complete overlooked possibility, not only in Denmark, but internationally. In pilot study, reference intervals has been established for UCBG and showed that arterial (from the neonate to the placenta) UCBG below 3.5 mmol/L (65 mg/dL) had an excellent diagnostic performance in predicting pathological HI in the four included cases.
The overall aim of this study is to further develop this new and proposed ideal screening tool, in diagnosing neonatal pathological HI with the perspective of global implementation.
Research Plan Study 1 and 4 are not included in this description.
Sub study 2. Prospective national study of UCBG screening for pathological HI Aim: to test the diagnostic performance of the pilot UCBG screening program on neonatal pathological HI in a prospective national 1-year cohort in Denmark.
Research question: Can measurement of arterial umbilical cord blood glucose (UCBG) with a cut-off value of ≤ 3.5 mmol/L (65 mg/dL) be used as an effective screening method to identify neonates with pathological hyperinsulinism (HI) in a national prospective birth cohort in Denmark? Methods: Expected approx. 9000 screen-positive cases among which a proportion (unknown, but estimated 30%) will have other risk factors ie. SGA, LGA or maternal diabetes, that would in any case have led to blood glucose monitoring.
- In collaboration with all Danish neonatal hospital units and their obstetrical wards, venous and arterial UCBG sampling will be optimized as a part of cord the routine gas analysis with implementation of routine UCBG data transfer to the hospital central labs.
- The pilot study normal reference for venous and arterial UCBG will be validated.
- In a smaller sample, variations of UCBG will be analyzed for the impact of time from birth to UCBG analysis. Expected 50 cases.
- Neonates with UCBG arterial glucose ≤ 3.5 mmol/L will be allocated to bedside glucose controls 1-2 and 3-4 hours after birth
- For determination of the false negative rate, a small sample of neonates (n=120) with UCBG arterial glucose above or at 3.5 mmol/L without any predefined risk factors of hypoglycemia, will undergo blood glucose measurements 6 times until 72 hours of age.
- Cases with pathological HI will be identified according to specified and uniform diagnostic criteria: Diagnostic test for hyperinsulinism (plasma glucose, serum insulin and C-peptide) during hypoglycemia will performed on the children according to a specified protocol, e.g. requirement of iv-glucose beyond 48 hours of age to prevent hypoglycemia; an i.v. glucose demand above 8 mg/kg/minute to obtain normoglycemia; prior to a decision of diazoxide treatment.
- Other information will be collected, including adherence to the new screening program, age at diagnosis of neonatal HI, severity of hypoglycemia and treatment.
- Concomitant bedside samples of glucose and ketones (β-OH-butyrate; together named G-BOHB) will be performed regularly until 72 hours' age or until intervention for hypoglycemia is started in cases with arterial UCBG <2.5 mmol/L and in the group of screen-negative patients (n=120) to evaluate the natural history of G-BOHB in the neonatal transitional hypoglycemia period (6 samples per neonate). Expected 50-100 cases + 120 screen-negative.
- For the perspective of global implementation, hand-held point-of-care (POC) glucometer UCBG will be measured in addition to cord gas glucose to evaluate the diagnostic performance of UCBG screening by POC glucometers.
Expected findings: Given the prevalence of transitional and persistent pathological HI and an annual birth rate of 60,000 in Denmark, a high participation rate and a diagnostic sensitivity close to 1,00, UCBG is expected to identify 50 Danish neonates with pathological HI during 1 year's national study. A high diagnostic performance and demonstration of robustness of the method will add to the fulfilment of optimal screening method criteria.
Inclusion: All live-born neonates delivered at Danish hospitals during the study period are eligible for inclusion. Exclusion: There are no predefined clinical exclusion criteria, but neonates for whom umbilical cord blood glucose measurement is not available due to unsuccessful or missing cord blood sampling will not be included.
Sub study 3. Neurological follow-up on patients with congenital hyperinsulinism after implementation of arterial UCBG screening
Aim: To evaluate the neurological and developmental status of children at the age of 6 months, diagnosed with pathological neonatal HI by UCBG screening. Research question: What is the neurological and developmental status at 6 months of age in infants diagnosed with pathological neonatal hyperinsulinism through umbilical cord blood glucose (UCBG) screening?
Methods: All pathological HI cases identified in Sub study 2 will be invited at 6 months' age for;
- Bayley Scales of Infant and Toddler Development IV: To assess fine and gross and motor development, giving each infant a raw and scaled score and percentile compared to the normal age-matched material.
- EEG: To detect eventual epileptiform activity as a complication to hypoglycemia.
- MRI scan: To identify cerebral complications to neonatal hypoglycemia.
Expected findings: Following the implementation of the new screening protocol for pathological HI, a significant reduction in the incidence of hypoglycemia-related NDI among affected infants is expected. As the study study has no control group, only literature data on the prevalence of NDI from congenital hyperinsulinism will be used for comparison. During the study period, two national laboratory data extractions of insulin measurements in newborns under 3 months of age will be done. This will serve two purposes: first, to ensure that all cases eligible for inclusion in Sub-study 3 are captured; and second, to assess adherence to UCBG screening and identify any false-negative cases. Insulin is only measured in newborns, when HI is suspected.
In the event of incidential findings, the child will be referred to a relevant hospital department for further evaluation and follow-up. Parents will be informed without delay, if findings requiring further assessment are identified. In cases where sedation is not sufficient, an individual assessment will be made in collaboration with the parents as to whether the MRI should be performed under general anesthesia. General anaesthesia will only be used if it seems clinically relevant and will be administered in accordance with standard paediatric anaesthetic guidelines.
Statistical analysis Diagnostic accuracy will be assessed by calculating sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and corresponding false positive and false negative rates. Estimates will be presented with 95% confidence intervals.
Receiver operating characteristic (ROC) curve analyses will be performed to evaluate the discriminative ability of arterial UCBG. The area under the ROC curve (AUC) with 95% confidence intervals will be calculated.
Neurological and developmental outcomes at 6 months of age will be analysed descriptively.
All analyses will be performed using STATA.
Perspectives Given the pilot study, it is expected that UCBG screening as an ideal screening tool for pathological HI will be confirmed with a low prevalence of NDI at 6 months' follow-up. This will significantly reduce the burden for the individual families and lead to reduction in societal costs of supporting and treating children with NDI. Demonstration of the usefulness of UCBG screening will provide the basis for an implementation in clinical practice, not only in Denmark, but world-wide implementation in collaboration with the Core Research Network under the global patient organization Congenital Hyperinsulinism International.
