Combined sulphonylurea and insulin treatment for type 2 diabetes mellitus : metabolic and electrophysiological studies

Abstract

Type 2 diabetes is a prevalent disorder characterized by elevated blood glucose levels and associated with increased morbidity from micro- and macrovascular complications. Lowering of blood glucose has been shown to reduce the incidence of such complications. Since the patient population is heterogeneous and the disease progresses with time, multiple therapies become necessary in the long-term. Combining sulphonylurea (SU) with insulin is one option, and various aspects of such therapy were studied in this thesis. Study I was a double-blind randomized study where 175 patients with secondary failure to oral agents were started on insulin in addition to oral glibenclamide. Glycaemic control improved, mean HbA1c 9.65% to 7.23% at four months. A high HbA1c and preserved insulin sensitivity at baseline were associated with a greater response to insulin combined with SU. Thereafter SU was, replaced by placebo in a majority of the patients, while a smaller control group continued on SU. The placebo/SU withdrawal group, fasting blood glucose (FBG) increased >10% in 79% within weeks, and 60% of the group were classified as SU responders, defined as >40% increase in FBG. Multivariate logistic regression analyses showed no clinically useful patient characteristics that identified SU responders at baseline, but long diabetes duration and absence of GAD antibodies were associated with beneficial response. A short period of SU withdrawal appeared as a useful test to determine whether a patient still benefits from SU.

In Study II 80 patients with secondary SU failure were randomized to two principally different insulin regimens - bedtime NPH or preprandial regular insulin - in addition to SU. Both regimens had similar effects on HbA1c but weight gain was more pronounced in patients on pre-prandial insulin. Patients who participated in Studies I and II at Danderyd Hospital were included in a follow-up over a median of 69 months (Study III). During this time four patients died and six were reclassified as having type I diabetes. Glycaemic control was maintained at an improved level with mean HbA1c 7.4% (SD 1.1) and weight stabilized after an initial gain of approximately 5 kg the first year. When necessary metformin was added to SU and insulin, and at 54 months about half were on such triple therapy. Hypoglycaemia is a limiting factor in attaining glycaemic goals in diabetes. Physiological responses to low blood glucose include decrease of insulin secretion and increase of counterregulatory hormones, the most important being glucagon and adrenaline. Study IV showed that glibenclamide affected these responses during insulin-induced hypoglycaemic clamp experiments in 13 patients with type 2 diabetes. When SU was present, insulin secretion was less suppressed and the glucagon response to hypoglycaemia was blunted. In Study V measurements of QT intervals were used to study the effects of hypoglycaemia on cardiac repolarization in 13 patients. Mean QT intervals and QT dispersion increased significantly during hypoglycaemia, indicating an increased risk of arrhythmia at low blood glucose values. No significant changes were seen between the two experiments with or without glibenclamide but the number of patients was too small to give conclusive data regarding this issue.

In summary, more studies regarding SU effects on glucagon secretion and cardiac repolarization are needed and should include comparisons of different SU derivatives. The present studies provide both pros and cons for combining glibenclamide with insulin in patients with type 2 diabetes who no longer can attain glycaemic goals on oral therapy alone; but the advantages seem to outweigh the possible disadvantages.