Clinical Utility of the T1D Genetic Risk Score—Examples from the U.S. Monogenic Diabetes Registry

Monogenic forms of diabetes are often difficult to distinguish from type 1 (T1D) and type 2 diabetes due to overlapping clinical features. As a result, a growing majority of patients remain misdiagnosed and often inappropriately treated. We used a T1D genetic risk score (T1D-GRS), generated from 10 T1D-associated common genetic variants, to help identify patients with a low T1D-GRS, and thus a potential monogenic etiology. We genotyped 798 probands from the U.S. Monogenic Diabetes Registry: 37% (n=293) had a known monogenic cause and 63% (n=505) had an unknown cause or were not exhaustively sequenced. To improve the discriminatory power of the T1D-GRS, participants were categorized using a T1D-GRS equivalent to the 5th, 25th, 50th, 75th, and 95th centile of the UK T1D cohort. 45% (132) of the registry participants with a confirmed monogenic cause had a T1D-GRS below the 5^th percentile, 35% (103) had a T1D-GRS between the 5-25^th percentile, 12% (35) between the 25-50^th percentile, 5% (15) between the 50-75^th percentile, and 2% (7) between the 75-95^th percentile. Of the participants with an unknown cause, 24% (n=108) had a T1D-GRS below the 5^th percentile, 30% (n=137) had a T1D-GRS between the 5-25^th percentile, 16% (72) between the 25-50^th percentile, 15% (67) between the 50-75^th percentile, and 15% (69) between the 75-95^th percentile. We sequenced a total of 124 participants with a T1D-GRS below the 25^th percentile using a targeted next-generation sequencing library that includes more than 140 genes known to be associated with diabetes. 21% (n=24) were found to have a pathogenic variant in a known diabetes gene. Further tests are ongoing. Our results show that the T1D-GRS is a useful discriminatory tool between diabetes types and a strong predictor of whether a monogenic cause is likely. More importantly, patients with a low T1D-GRS and negative for all known monogenic subtypes should be prioritized for exome or whole genome sequencing to maximize the chance of novel gene discoveries.