Predicting Hypoglycaemia and Arrhythmias in the Patient With Diabetes and CKD - Validation Study

Patients with diabetes mellitus (DM), especially those with a long duration of diabetes, insulin treatment and chronic kidney disease (CKD) are vulnerable patients exhibiting a high risk for cardiac arrhythmias and sudden cardiac death (SCD). Various factors such as the presence of coronary heart disease, diabetic cardiomyopathy as well autonomic neuropathy are underlying pathologies associated with the development of potentially fatal arrhythmias in these patients while hypoglycemic events are considered to directly trigger these arrhythmias. It has been postulated that severe hypoglycemia may lead to cardiac arrhythmias, later summarized as the "dead in bed" syndrome. In addition, recent data from large cardiovascular outcome trials in patients with type 2 diabetes suggest that severe hypoglycemia is associated with an increased risk of cardiovascular events and cardiovascular related death. Moreover, CKD markedly increases the risk for hypoglycemia and even a moderate impairment of kidney function (eGFR < 45 ml/min) is associated with a significant increase in SCD.

Various pathophysiological mechanisms may contribute to the increased cardiovascular mortality after hypoglycemia including hypoglycemia-induced release of catecholamines, pro-arrhythmogenic ECG alterations as well as inflammatory changes. Morphological and functional alterations of the heart occurring in CKD further contribute to these mechanisms. So far performed studies were limited by a short duration of glucose and ECG monitoring and by the fact that only 3 lead Holter-ECGs were used, thus not allowing the assessment of more sophisticated ECG abnormalities such as QT dispersion, T-wave alternans, or late potentials. Therefore no clear data exist to predict arrhythmias and SCD and its relation to hypoglycemia in patients with diabetes. Ideally, a SCD risk score could identify and characterize high-risk patients but to date little is known about hypoglycemia-associated ECG markers for the identification of patients at risk for arrhythmias and SCD.

In the general population, various ECG risk markers for SCD have been identified such as heart rate, cardiac rhythm abnormalities, atrioventriculare (AV) block, QT length, QT dispersion, heart-rate variability (HRV), T-wave alternans, late potentials, as well as left- (LBBB) or right-bundle branch block (RBBB) (reviewed in). In patients with diabetes hypoglycemia, diabetic cardiomyopathy, as well as the presence of autonomic neuropathy may lead to such ECG abnormalities. Merely sparse data exist on the effect of spontaneous hypoglycemic episodes and changes in ECG parameters with only a small study in patients with type 1 diabetes demonstrating that nocturnal hypoglycemia is associated with a decrease in the low-frequency component of heart rate variability. To date, more sophisticated markers such as QT dispersion, late potentials, or T-wave alternans were not examined in a "real-life setting", most likely because these markers require a 12 lead ECG registration of longer duration.

However, for the establishment of a risk algorithm for the prediction of hypoglycemia-associated arrhythmias it is mandatory to perform long duration simultaneous glucose monitoring and 12 lead ECG registration to capture these ECG risk markers for SCD.

In an actually running project the investigators are evaluating the association of hypoglycemic events/glucose swings and arrhythmias/ECG predictors for SCD a clinical study will be performed in 50 patients with insulin-treated diabetes and moderate to severe CKD. These patients receive 7 days continuous glucose and ECG registration and data will be used for the development of the risk assessment model.

The current validation study seeks to confirm the risk assessment model developed in collaboration with AICES - Aachen Institute for Advanced Study in Computational Engineering Science and to approve the capacitive ECG registration device obtained in collaboration with Philips Chair for Medical Information Technology at University Clinical Center Aachen (UKA). To this end, 10 patients with insulin-treated diabetes and moderate to severe CKD will be included. Seven day glucose monitoring as well as data of capacitive ECG recordings will be generated in this study, thus allowing validation and adjustment of the developed medical hardware and the mathematic models. The study item is a 12-lead ECG T-shirt consisting of textile electrodes and a data logging device. The device can record long-term 12-lead ECG data. The purpose of the T-shirt is to improve the patient's comfort for long-term recordings and to prevent adverse effects of regular ECG electrodes. Current systems are limited by the use of ECG electrodes, which are hardly tolerated by the patients because of severe direct side effects on the skin such as rash and bullous lesions. These side effects are a result of skin preparation and electrode gel. The proposed ECG T-shirt does not need these problematic preparations. Another benefit is the fixed placement of the electrodes on the T-shirt. In regular 12-lead ECG long-term recordings, the electrodes may fall off and the patient needs to reattach them. Therefore, faulty positioning of the electrodes may occur.

Inclusion criteria

• Insulin-treated diabetes mellitus (type 1 or 2)
• CKD with eGFR < 45 ml/min, determined using the Modification of Diet in Renal Disease (MDRD) formula
• Stable anti-diabetic and cardiac medication prior to inclusion
• Male aged ≥ 18 years
• Written informed consent prior to study participation
• Adults who are contractually capable and mentally able to understand and follow the instructions of the study personnel.

Exclusion criteria

• Pregnancy or women without sufficient contraception, adapted specifically to amenorrhoeic hemodialysis patients
• Life expectancy below 6 months
• Participation in another clinical trial within the previous 2 months
• History of any other illness, which, in the opinion of the investigator, might pose an unacceptable risk when administering study medication
• Any current or past medical condition and/or required medication to treat a condition that could affect the evaluation of the study
• Alcohol or drug abuse
• Patient has been committed to an institution by legal or regulatory order
• Patients with any kind of pacemakers
• Expected non-compliance
• Patients unwilling or unable to give informed consent, or with limited ability to comply with instructions for this study
• Participation in a parallel interventional clinical trial