Researchers from the University of Exeter in the U.K. have shown that incorporating a type 1 diabetes genetic risk score (T1DGRS) into standard testing for maturity-onset diabetes of the young (MODY) could identify hidden type 1 diabetes in around one in six patients with unresolved test results.
“In clinic, it’s extremely challenging to distinguish between MODY and type 1 diabetes, yet getting the right diagnosis is crucial to getting the right treatment,” said lead author Kashyap Patel, associate professor at the University of Exeter, and consultant physician in Diabetes and Endocrinology. “We found that applying the T1DGRS to patients who tested negative for MODY effectively identifies individuals with type 1 diabetes, reducing the need for expensive further testing and providing answers to patients.”
MODY is a familial form of diabetes that accounts for approximately 3.6% of diabetes diagnoses before age 30 years. However, due to overlapping clinical presentation with type 1 diabetes, around 80% of individuals with MODY are initially misclassified.
Patel and team therefore evaluated the clinical utility of incorporating the T1DGRS into routine MODY gene panel testing among 1129 individuals treated with insulin and referred for MODY genetic testing. Of these, 20.2% had MODY-causing variants identified.
Following gene panel testing that included 10 variants associated with type 1 diabetes, the researchers calculated that participants had a median T1DGRS of 0.58 points, on a scale of 0 to 1. By comparison, the median scores were 0.55 points in a reference group of individuals with non-autoimmune diabetes and 0.69 points in a type 1 diabetes reference group.
“This intermediate distribution indicates that the referral cohort comprises a mixture of [type 1 diabetes] and nonautoimmune diabetes,” Patel and co-authors write in Diabetes Care.
By modeling the T1DGRS distribution against reference populations, the researchers estimated that 20.0% of the insulin-treated referral cohort had atypical type 1 diabetes. This proportion increased to 26.0% among the patients with unsolved MODY.
When the genetically unsolved MODY cohort was stratified by age, the T1DGRS was highest in children, with estimated type 1 diabetes prevalence at 76.2%, 25.5%, and 16.8% among those aged below 10 years, 10–18 years, and over 18 years, respectively.
Using this information, the investigators created age-specific T1DGRS thresholds, which when applied identified type 1 diabetes in 16.1% of unsolved MODY cases.
Finally, the team assessed the discriminatory performance T1DGRS as a pre-genetic testing tool for identifying individuals with a low likelihood of MODY.
In the overall cohort, T1DGRS showed limited discrimination with an area under the curve (AUC) of 0.60 but the test performed well in children aged below 10 years, at an AUC of 0.83. At the optimal threshold, no MODY diagnoses were missed in the unsolved MODY group and 53.2% of individuals could have been excluded from MODY testing.
“These findings suggest that although the T1DGRS is not suitable as a universal prescreening tool, its improved performance in children may support further evaluation of its use in age-specific MODY referral pathways,” Patel et al. conclude.
The approach will now be adopted as part of NHS testing, said co-author Kevin Colclough, lead clinical scientist for the Exeter MODY testing service at the Royal Devon University Healthcare NHS Foundation Trust. “This will be a significant quality improvement to our service, providing clinicians with a more informative result and helping to reduce diagnostic uncertainty for their patients,” he remarked.
