Delayed hypoglycemia induced by repaglinide in a frail elderly adult with diabetes mellitus.

Abstract

To the Editor: Type 2 diabetes mellitus has a 20% prevalence in individuals aged 75 and older in industrialized countries. In France, three-quarters of individuals with diabetes mellitus are treated with oral hypoglycemic agents: mainly metformin (69%), sulfonylureas (40%), and repaglinide (21%).1 The risk of hypoglycemia induced by hypoglycemic agents has been evaluated in studies conducted in young and vigorous elderly adults, but frail elderly adults have been excluded from these studies. Repaglinide is the only glinide available in France. Clinical studies investigating the use of repaglinide in individuals aged 75 and older are lacking, which is a problem because it widely prescribed in this population. A frail 84-year-old man living alone in the community had been hospitalized for delirium with hyperactive and hypoactive episodes 2 weeks earlier. He had had type 2 diabetes mellitus for 20 years and had heart failure and moderate to severe cognitive decline. A nurse came daily to administer medication. His weight was 76 kg (body mass index 25 kg/m2). Blood parameters were hemoglobin 114 g/L, fasting glucose 126 mg/dL (7 mmol/L), glycosylated hemoglobin (HbA1c) 7.4%, creatinine 77 μmol/L (glomerular filtration rate 89 mL/min per 1.73 m2 according to the Modified Diet in Renal Disease), and albumin 42 g/L. His medications were furosemide (40 mg/d), esomeprazole (20 mg/d), acetylsalicylic acid (75 mg), nicorandil (20 mg/d), diltiazem (240 mg/d), amiodarone (200 mg/d), ebastine (10 mg/d), escitalopram (5 mg/d), metformin (1,000 mg twice daily), and repaglinide (4 mg at breakfast, 2 mg at lunch, and 4 mg at lunch). Continuous glucose recording (CGR; Medtronic, Northridge, CA) was initiated because of the discrepancy between the HbA1c value that indicates good glucose control (7.4%) and high finger-stick glucose measurements (FSGMs). CGR showed glucose levels of <70 mg/dL from 2:00 a.m. to 6:00 a.m. and, to a lesser extent, from 4:00 p.m. to 7:00 p.m. (Figure 1A) without clinical symptoms, whereas FSGM (8/day including nocturnal measurements) did not detect the hypoglycemic episodes. At this time, his Mini-Mental State Examination (MMSE) score was 15 out of 29. Repaglinide was not withdrawn but was rapidly reduced until a dose of 0.5 mg at breakfast and dinner was reached. There was not a dramatic increase in FSGM during postprandial periods. Three months later a CGR was performed during a short hospitalization (Figure 1B). At this time, his MMSE score was 20 out of 29. Repaglinide lowers blood glucose by enhancing insulin release from pancreatic β-cells and prevents postprandial hyperglycemia. The frequency of symptomatic hypoglycemic episodes (plasma glucose <70 mg/dL (4 mmol/L)) is lower with repaglinide than glibenclamide (a sulfonylurea).2 Repaglinide is reported to be well tolerated in elderly adults because of its short plasmatic half-life (<1 hour), metabolism in the liver by CYP3A4 (one of the cytochrome P450 isoforms without active metabolite release), main excretion through the biliary route, and low renal excretion (8%).3 Individuals with moderate renal impairment may be treated without initial dose adjustments of repaglinide.3 Drugs that increase or decrease levels of CYP3A4 in the liver may affect blood levels of repaglinide and thus alter its glucose-lowering effect, but this man had not been given such a medication. FSGM frequently does not detect hypoglycemic episodes in older adults.4 The CGR highlights hypoglycemic episodes that FSGM does not detect.4, 5 Repaglinide controls postprandial hyperglycemia, but that was not the case here, with postprandial glycemia reaching 300 mg/dL, although seemed to act as a long-acting insulin-releasing drug responsible for nocturnal hypoglycemia several hours after its last administration. Fasting plasma glucose, capillary glucose (8/day), and HbA1c measurements did not detect this condition. At the same time, repaglinide was not effective at controlling postprandial hyperglycemia—up to 300 mg/dL—as had been expected (Figure 1A). Postprandial blood sugar did not increase after dose reduction. Finally, chronic hypoglycemia might exacerbate or be responsible for cognitive and depressive symptoms, and CGR might be a useful tool to explore cognitive disorders in older adults with diabetes mellitus who are given hypoglycemic agents. Conflict of Interest: Mennecart, Constans: Diabéto Centre meeting, Sanofi, Tours, 2013. Constans: Novo Nordisk symposium on hypoglycaemia, SNFMI congress, Poitiers, 2011; EASD congress, Barcelona, 2013 (BoehringerIngelheim) and Journées de Diabétologie, Paris, 2014 (Lilly). Mennecart: Agediab, Novo Nordisk, Tours, 2013. Author Contributions: Mennecart, Constans: interpretation of data, drafting of the manuscript, final approval of version to be submitted. Malherbe: use of glucose recording device, acquisition of data. Mondon: evaluation of cognitive disorders, critical revision of manuscript. Sponsor's Role: None.