Jan Ruxer

Discontinuing insulin therapy after diabetic ketoacidosis—is its cause worth considering?

We read with interest the paper by Hsin Yu et al . [1], and found it of significant practical importance. However, we believe that the authors missed an important issue in clinical characteristics of diabetic ketoacidosis (DKA), as no data on possible causes of DKA in the studied group are presented. DKA might occur in diabetes at the onset of the disease, and later on due to insulin omission or infection [2,3]. It is less clear what causes DKA in newly diagnosed subjects, who later do not require insulin to control blood glucose. We suggest that a specific cause of DKA might also be a useful predictor of insulin discontinuation in further stages of the disease. We describe a case of a patient who was diagnosed with diabetes and DKA, in whom insulin was withdrawn after 2 months. He presented with all three factors associated with insulin discontinuation described by Hsin Yu et al . [1]. On 1 February 2001, the patient (male, Caucasian, aged 53 years, body mass index 28.7 kg/m 2 ) was referred to the Metabolic Diseases Department with DKA and an upper respiratory tract infection. His personal medical history was unremarkable, although three members of the patient’s family were diabetic. In the previous 3 days he had noticed increased thirst and polyuria. He presented with features of diabetic ketotic acidosis (blood glucose 41.1 mmol/ l, pH 7.26, serum bicarbonate 9.8 mmol/ l, base excess [ − 14.9] mmol/ l), was dehydrated, tachypnoeic, and had symptoms of severe purulent pharyngitis. Other laboratory abnormalities included elevated leucocyte count (14 850/ μ l, with 83.7% of neutrophils) and significantly elevated creatinine kinase (CK) up to 50 880 IU/l (reference range 10–90 IU/l). Intravenous insulin as well as antibiotic and fluids were initiated and the patient’s general condition improved promptly. CK levels decreased to normal values within 6 days as did his blood cell count. His initial daily insulin dose was 103 IU, after 7 days it decreased to 72 IU, and at discharge, after 15 days of hospital treatment, he was taking 52 IU of insulin per day in multiple doses. His anti-GAD assay was negative. Ophthalmologic and neurological examinations were unremarkable. After hospital stay the patient’s insulin requirement was decreasing steadily, and after 6 weeks insulin therapy was discontinued. He was switched to glibenclamide 5 mg t.i.d. Nevertheless, he experienced hypoglycaemia and eventually, after 4 months, all pharmacological treatment was stopped. His mean daily blood glucose was between 4.5 and 5.0 mmol/l and his diabetes was managed by diet. The case is a good example of a person in whom insulin therapy discontinuation would be expected according to the criteria proposed by Hsin Yu et al . [1]. He was over 40 years of age at diagnosis, DKA was his first symptom of diabetes and he was overweight. However, we believe that the cause of DKA should also be taken into consideration. Severe infection is a wellknown precipitating factor of ketoacidosis in diabetic patients [2]. Our patient developed DKA following severe pharyngitis, but within weeks of its treatment no longer needed insulin. It would therefore be interesting to learn which subjects studied by Hsin Yu et al . presented with infection. This might be another factor predicting insulin discontinuation. Finally, the general conclusion drawn from the work of Hsin Yu et al . and other observations [4,5] suggests that ceasing insulin therapy is not uncommon in diabetes patients. However, more studies are needed to shed light on this somewhat understudied area of insulin therapy.