Providing information for patients and professionals on research and clinical care in genetic types of diabetes.

Genetic Testing for Neonatal Diabetes

A genetic test can alter treatment, predict prognosis and improve clinical management.

The genetic team lead by Professors Andrew Hattersley and Sian Ellard in Exeter identified activating mutations in the KCNJ11 and ABCC8 genes, encoding the Kir6.2 and SUR1 subunits of the pancreatic K-ATP channel as the most common cause of neonatal diabetes which is typically diagnosed in the first six months of life (Gloyn et al. N Eng J Med 2004:350; 1838-1849; Proks et al, Hum Mol Genet 2006:15; 1793-1800). These patients often have de-novo mutations and so do not have diabetic relatives. However, around 10% show generation to generation transmission with an autosomal dominant inheritance and about 20% of patients have neurological features, typically developmental delay and in some cases epilepsy (termed DEND syndrome). Beta-cell autoantibodies seen in Type 1 diabetes are not detected. Patients usually present with very high glucose values and occasionally ketoacidosis and often C peptide is not measurable even in response to either glucose or glucagon. Treatment with high dose sulphonylureas is successful for most patients and achieves improved glycaemic control. Protocols for transfer and further information can be accessed here shortly «link pending».

To date, at least 300 patients with a Kir6.2 or SUR1 mutation have been successfully treated with oral sulphonylurea tablets (stopping insulin completely with better control and less hypos) despite being apparently insulin dependent. Reports of this include:

1. Klupa T, et al. Diabetologia. 2005; 48(5):1029-31
2. Codner E, et al. Diabetes Care. 2005; 28(3)758-9
3. Zung A. J Clin Endocrinol Metab. 2004; 89(11):5504-7
4. Sagen JV, Diabetes. 2004; 53(10):2713-8
5. Pearson E, N Eng J Med 2006; 355(5): 467-477
6. Rafiq M, Diabetes Care 2008; 31(2): 204-209

A minority of patients have not managed to transfer from insulin. Most of these have had very severe neurological features with profound developmental delay and generalised epilepsy before 6 months or were adults when the change of therapy was introduced.

Whilst the majority of patients with a Kir6.2 or SUR1 mutation are diagnosed with diabetes before 6 months of age there are 3 reports where mutations have been identified in infants with diabetes diagnosed between the ages of 6 and 12 months (Mohamadi et al Pediatr Diabetes, 2010. 11: 203-7; Rubio-Cabezas et al, Pediatr Diabetes, 2012. 13: 322-5).

Which patients should be tested?

We provide diagnostic testing for all patients diagnosed with diabetes under the age of 6 months whatever their age now. We will also screen for Kir6.2 and SUR1 mutations in patients diagnosed between the ages of 6 -9 months with antibody negative diabetes although the likelihood of identifying a mutation in these patients reduces from approximately 50% when diagnosed before 6 months of age to less than 5% (Rubio-Cabezas et al, Pediatr Diabetes, 2012. 13: 322-5). Genetic testing may also be provided for patients diagnosed with diabetes between 9-12 months who are clinically atypical however each patient will be considered for testing on an individual basis. Please contact Prof Sian Ellard ( to discuss these cases.

Specific features that increase the chance of diagnosis are: low birth weight 1.8 - 3.1kg at term, a family history of neonatal diabetes or marked developmental delay, but the latter 2 features are only seen in a few cases. The present age of the patient is not important and patients diagnosed in the first six months of life who are now under the adult physicians would be appropriate for testing.

We would strongly recommend taking blood or DNA from both parents as well as the child, as this will allow us to establish rapidly that a genetic mutation is likely to be the cause of the diabetes if it has arisen de novo and is not present in either unaffected parent.


Genetic testing for neonatal diabetes is provided free of charge for all patients until 2020.  This has been made possible by funding from the Wellcome Trust in the form of a Senior Investigator Award to Professors Hattersley and Ellard.

Details to send with samples

Please when you send us a sample for genetic testing; please send us all details on our request form. These include:

  • Referring clinician details
  • Patient Details
  • General
  • Presentation
  • Current status of Patient
  • Any other clinical features


When taking samples you should get written consent for the testing - as with any genetic test. You can download the consent form that you might like to use, modify, or translate. We ask you get consent from the patient/parents to test for the genetic cause of the patient's diabetes rather than limiting to Kir6.2 and SUR1.

Samples to be taken

The samples should be labelled with name and date of birth and can either be:

1. Our preferred option is Blood 3-5mls: taken in tubes containing EDTA and couriered fresh (not frozen). Must arrive in UK within 5 days, so ideally send on a Monday. This should be sent in a package with at least two layers of wrapping and include absorbent material to absorb any blood that leaks out of the tube.
2. Or made into DNA in your own country and 2-10 micrograms (to allow repeats) sent at room temperature. Again please make sure the tube is very securely sealed.

Where should the sample be sent?

Please e-mail the patient’s name and date of birth to together with the completed electronic request form and/or additional clinical details. The details will be pre-logged so that when the samples arrive we can match up the information and send you an e-mail confirming sample receipt.

The sample should be sent to:

Prof. Sian Ellard PhD FRCPath
Department of Molecular Genetics

RILD Level 3
Royal Devon and Exeter NHS Healthcare Trust
Barrack Road
Exeter, EX2 5DW


We treat these samples according to clinical diagnostic testing requirements and hence do use names and dates of birth to give us unique identifiers for each sample. This information is then dealt with in complete confidentiality. This means we issue a report with the patient's name and date of birth on it that can join their clinical records while an anonymous one could not. Using the patients' names also prevents the same patient being referred by more than one clinician. We would request that a sample comes with a name and date of birth so it can be logged onto our IT system and we can always find the sequence etc. if there are any queries. Our lab is dealing with approximately 80 diagnostic DNA samples a day and full confidentiality has to be maintained for all of these so this is not an issue.


We will send a result as soon as possible, usually by email. Samples from newly diagnosed babies are prioritised for testing and a result is often available within 1 to 2 weeks. All physicians referring patients will receive a written report 6-8 weeks after the receipt of the sample. With mutations that have not previously been described, it is hard to be certain if they are pathogenic. This would be established if the mutation is not present in unaffected parents or by co-segregation with neonatal diabetes in families with multiple affected members. This is why we are keen to have DNA from both parents with the initial sample sent.

Other genes

If we do not find a Kir6.2 or SUR1 mutation we will test other known and possible causes of neonatal diabetes and send a further report by e-mail 2-3 months later. Testing of the known genetic causes by targeted next generation sequencing identifies a mutation in ~80% of patients. Further testing by exome or genome sequencing is undertaken in selected cases according to phenotype and availability of parental samples.

Download ISPAD Guidelines on the diagnosis and management of monogenic diabetes in children

Other questions

Please do contact us if there are any other questions. We would be very keen to help in any way we can.

  Professor Andrew Hattersley Professor Sian Ellard
Tel: +44 (0) 1392 408260 +44 (0) 1392 408259