6410 Investigating possible genotype-phenotype relationship in patients with neonatal diabetes caused by mutations in GATA6 and PDX1

Objectives Introduction: Neonatal Diabetes Mellitus (NDM) is a rare genetic condition (~ 1 in 90,000 live births) with over 30 known genetic causes occurring mainly before 6 months of age. It is defined by the presence of severe hyperglycaemia associated with insufficient or no circulating insulin. Variable phenotypic spectrum has been observed in NDM patients with different genetic subtypes, including those with mutations in either GATA6 or PDX1 genes, exhibiting complex cardiac defects, pancreatic, and gall bladder malformations and other abnormalities. Aim: To underscore the phenotypic variability of individuals with mutations in GATA6 or PDX1 in order to determine whether there is a phenotype-genotype correlation. Methods We classified 95 variants in the GATA6 gene from 148 patients and 22 variants in PDX1 from 37 patients. Variants were classified according to the ACGS Best Practice Guidelines, and clinical data was obtained from published literature and referral notes for individuals referred to the Exeter Genomics Laboratory. The Chi-square test was used to investigate phenotype-genotype relationships. Results 66 GATA6 variants were identified as likely pathogenic/pathogenic in 107 individuals and are categorised into 22 missense variants, 9 splicings, 23 insertions/deletions (frameshifts) and 12 non-sense changes. These patients either had diabetes with an extra-pancreatic phenotype (n=66) or extra-pancreatic characteristics without diabetes (n=41). We found that de novo mutations (p-value <0.0001), missense changes in the second GATA6 Zinc Finger (ZF) binding domain (p=0.014), and loss- off-function GATA6 variations (p=0.003) were more likely to result in patients having diabetes. Among 26 individuals with recessive PDX1 variants, 19 disease-causing variants were identified which include 15 missense variants, 3 frameshifts, and 1 non-sense. These individuals either had a pancreatic phenotype only (NDM with or without exocrine pancreatic insufficiency) or a pancreatic phenotype with extra-pancreatic features. The genotype-phenotype correlation within this cohort was not significant. However, our findings showed that 1 in 3 patients with PDX1 mutation present additional features, affecting the gut or liver along with NDM. Conclusion Our study confirmed a statistically significant phenotype-genotype relationship in individuals with pathogenic GATA6 mutations. These findings might increase clinicians' awareness of the possibilities and implications of GATA6 and/or PDX1 mutations in individuals with diabetes and heterogeneous clinical features.