Continuous glucose monitoring - proved hypoglycemia in sodium-glucose co-transporter-2 inhibitors - treated type 2 diabetes patients: a link to ketosis development

Abstract

Background: Latest studies have shown the remarkable ability of sodium-glucose co-transporter-2 inhibitors (SGLT2i) to reduce cardiovascular morbidity and mortality. However, real-life data and the results of several other studies seem to contradict these outcomes, pointing out possibilities of serious side effects. Ketoacidosis (KA) remains one of the most dangerous complications, yet, not fully understood. All of the above urgently requires real-practice data, which may shed some light on side effects of this novel anti-diabetic drug family.
 Aims: To investigate the real-life rates of hypoglycemia and ketosis (K) in SGLT2i treated patients, using Continuous Glucose Monitoring (CGM) and capillary blood -hydroxybutyrate measurements.
 Methods: We report the results of a two-year retrospective analysis of 136 Type 2 Diabetes (T2DM) patients, all (100%) treated with a SGLT2i, combined with Metformin or Metformin with Incretin-Based therapy (MT-IBT). CGM recordings were done in 52 persons. In 9 patients (Group A), CGM-proved hypoglycemic episodes were discovered. The rest of 43 patients (Group B) didnt show any hypoglycemia. Three patients in Group A and 11 from Group B were also treated with small doses of basal insulin on admission; the insulin was later discontinued in all patients of Group A and seven patients of Group B . Main characteristics of two groups were subsequently compared.
 Results: CGM data analysis showed significantly lower average Sensor Glucose (SG) , 7.21.3 vs. 8.21.7 mmol/l, p=0.04, and estimated HbA1c , 6.10.7 vs. 6.81.1%, p=0.02, in Group A patients.
 We also report three cases of ketosis, proved by elevated capillary -hydroxybutyrate concentrations. Pathophysiological link between frequent hypoglycemia rates (Six patients without insulin treatment (11.5 % in total CGM group of 52 patients)) and ketosis development (Three patients (2.2% in total cohort of 136 participants)) was suggested.
 Conclusions: More frequent than previously reported rates of hypoglycemia and ketoacidosis were discovered in patients taking SGLT2 inhibitors. Pathophysiological link between the two conditions is assumed. More studies are needed to confirm our hypothesis.