Comment on: Khurana et al. The diagnosis of neonatal diabetes in a mother at 25 years of age. Diabetes Care 2012;35:e59.

We read with interest the observation by Khurana et al. (1) regarding the speculative treatment of their case of neonatal diabetes (ND) pending molecular genetic confirmation. The authors report the successful trial of a sulfonylurea in a baby diagnosed with diabetes on day 2 of life in whom an activating KCNJ11 mutation was subsequently identified. Khurana et al. suggest a controlled trial of sulfonylurea in patients with ND if genetic testing is not feasible.

We would not agree with this suggested management plan for two reasons: Firstly, not all patients with ND will respond to treatment with a sulfonylurea, and secondly, the result of genetic testing for ND can be available in a week.

Of the known genetic subtypes of ND, only patients with an activating KCNJ11 or ABCC8 mutation will respond to treatment with a sulfonylurea (2,3). Within our international cohort of >1,000 patients from 75 countries with ND diagnosed before the age of 6 months, 338 were diagnosed within the first week of life. Of these, 34 have a confirmed ABCC8 mutation and 31 have a confirmed KCNJ11 mutation. Thus only 19% of those diagnosed with ND in the first week of life have a high probability of responding to treatment with a sulfonylurea. This does not exclude the possibility of other as yet unidentified genetic causes of sulfonylurea-sensitive ND, but to our knowledge there are no such reported cases.

Rapid sequence analysis of the KCNJ11 or ABCC8 genes is possible within a week of receiving a sample. This service is provided by our laboratory without charge (funded by the Wellcome Trust) for patients throughout the world who are diagnosed with diabetes within the first 6 months of life (www.diabetesgenes.org). Other laboratories offer a similar service. Given the speed of receiving a genetic result and that response to sulfonylurea is only likely in approximately 1 in 5 of ND cases diagnosed within the first week of life, we strongly recommend continuing insulin therapy until a genetic diagnosis is available.

Treatment of a patient with a sulfonylurea without a molecular diagnosis is hard to justify. Insulin treatment to control blood glucose until a diagnosis is made will mean the child will benefit from the anabolic effects of insulin as well as the correction of hyperglycemia. This treatment will not harm the neonate, nor affect long-term outcome—and will ensure that blood glucose is well-treated during this critical neonatal period. Conversely, speculative treatment with a sulfonylurea is dangerous as it can mean a patient without a potassium channel mutation has no treatment for hyperglycemia but also that clinicians may stop the therapeutic trial inappropriately early as some patients need prolonged exposure to high-dose sulfonylurea before they respond.

The treatment, clinical course, prognosis, and associated features will all vary with the molecular genetic etiology of a patient’s ND. We advocate the recommendations of the International Society for Pediatric and Adolescent Diabetes (4) that blood be sent for a molecular genetic diagnosis as soon as the diagnosis of diabetes is made. We hope our free, rapid, international diagnostic service will make this possible.