P324: Research analysis of clinical exome sequence data to detect diabetes-related genetic variation

Introduction: Monogenic diabetes (MODY) accounts for 1-4% of all diabetes cases in children and adolescents.Identification of molecular causes of MODY is important to ensure appropriate patient treatment and genetic counseling.In this study, we sought to investigate the diagnostic yield of five MODY-related multigene panels in 80 patients who previously received exome-based gene panel (EBGP) testing in the Clinical Genomics Laboratory (CGL) at Washington University School of Medicine.To this end, we conducted reflexive research analysis of previously generated exome sequencing (ES) data to determine whether it may increase diagnostic yield and/or lead to the discovery of biologically suspicious variants in genes not yet associated with MODY.Methods: Previously generated ES data were processed and annotated using the Franklin analysis engine by Genoox.Curated variants were categorized as follows.1) Variants identified in previously assessed EBGP genes; 2) variants in genes associated with diabetes but not included on the previous version of the EBGP; 3) variants in candidate disease genes (or genes of uncertain significance (GUS)); and 4) variants in genes associated with an initially untargeted phenotype.The potential relevance of GUS variants was assessed based on published literature, gene constraint metrics, and bioinformatic predictions.Patient clinical information, including detailed phenotypes, was aggregated from electronic health records.Results: The MODY clinical testing strategy in the CGL includes use of five different EBGPs, encompassing a total of 54 unique genes.EBGPs were curated based on known diabetes-gene associations, and phenotype specific panels were selected based on provider clinical suspicion (eg, hyperinsulinism, etc).To date (November 2024), the CGL has completed genetic testing for 80 endocrine patients with suspected MODY; most test orders (52/80 total tests) were for the monogenic diabetes and maturity onset diabetes of the young (MODY) panel.Nine percent (n=7/80) of previous clinically tested patients received a positive result, while 91% (n=73) received inconclusive or negative findings.ES data representing the 80 patients (45 under age 18) was analyzed to detect MODY-relevant variants.Initial ES analysis identified 146 variants requiring in-depth review, resulting in 25 candidate variants that were categorized as described in the methods.In short, no previous EBGP identified inconclusive findings were reclassified as diagnostic, and no additional diagnostic findings were detected via reflex ES analysis.Interestingly, we identified nine variants residing in candidate diabetes genes or genes not previously associated with MODY phenotypes, including two patients harboring different predicted deleterious missense variants in CAPN10 (associated with increased adiposity and palmitate-induced pancreatic apoptosis).All GUSs have been submitted to GeneMatcher.Further, we identified two well-established pathogenic variants in ARSA, associated with autosomal recessive metachromatic leukodystrophy, in a young patient who currently exhibits no accompanying disease features (variant phase is unknown).Conclusion: In our cohort of patients with diabetes, reflex research analysis of ES data resulted in a diagnostic yield comparable to traditional EBGP testing (ie, no additional diagnostic variants were detected).However, ES provided the added benefit of detecting variants in candidate GUSs that might eventually lead to the discovery of new diabetes associated genes.In addition, ES analysis identified potential clinically relevant variants in one patient not associated with the initial indication for testing.