276-OR: Common Variant Polygenic Risk Score Influences Penetrance of Rare Variants in Monogenic Diabetes Genes

Monogenic diabetes (MD) accounts for 0.4% of diabetes cases and is a paradigm for precision medicine. However, a high fraction of variants in MD genes reported in ClinVar have uncertain significance (VUS) or conflicting interpretations of pathogenicity (CIP), making potential MD diagnosis a challenge.We hypothesized that using large-scale rare variant association data for type 2 diabetes (T2D) risk could inform variant pathogenicity in the MD genes. We tested the association with T2D of 1,157 VUS/CIP in 52 established MD genes, in the largest rare-variant T2D genome-wide association study (51,256 T2D, 370,487 controls), using rare variant imputation and whole-genome sequencing. We identified 9 variants with intermediate penetrance (VIP) (odds ratio [OR] > 5) in HNF1A, HNF4A, GCK, KCNJ11, WFS1, and POLD1, including a missense variant in HNF4A (p.Arg114Trp) associated with an 8-fold increased risk of diabetes (OR=8.2, [4.6-14], p=1.1×10^-13). The penetrance of this variant was influenced by a T2D polygenic risk score (PRS), with carriers in the highest tertile of the PRS showing a risk (OR=18.3 [7.2-46.9], p=1.2×10^-9) comparable to the effect of confirmed pathogenic HNF4A MODY variants (OR=19.3 [6.1-60.6], p=4.1×10^-7), and carriers within the lower tertile of the PRS showing a much smaller risk (OR=2.6 [0.9,7.1], p=0.06). We also provided support for 21% of rare variants in the MD genes as benign based on their sufficient statistical power but null association with T2D. We show that large-scale association data can inform variant pathogenicity and that rare variant penetrance is influenced by PRS. Our findings support the existence of a spectrum between common and monogenic forms of diabetes rather than a categorical distinction.