ESPE Abstracts

Congenital hypopituitarism (CH) is a highly variable disorder affecting 1:3000 - 1:4000 live births, and is characterized by deficiencies in one or more of the 7 pituitary hormones, with growth hormone (GH) being the most frequently-occurring deficiency. It may be associated with a range of syndromic features including visual impairment, midline brain abnormalities and facial clefting. The development of the pituitary gland is closely associated with that of the forebrain and hypothalamus, and involves a cascade of transcription factors and signalling molecules, intrinsic and extrinsic to the developing anterior pituitary, essential for correct hypothalamo-pituitary (HP) development. We have collected DNA from a cohort of patients with CH and related disorders, followed up at Great Ormond Street Hospital (GOSH) and other national/international centres (n=>1900). Our previous studies have revealed mutations in known causative or novel genes in ~10% of patients, with no aetiology identified in up to 90%. We have set up a new targeted gene panel, including 135 genes, both known and novel, that have previously been associated with CH or related disorders, or that have been implicated as playing a role in murine or human HP development e.g. HESX1, GLI2, SOX2/3, LHX3/4 etc. We have incorporated genes from our exome sequencing projects (published and unpublished) e.g. EIF2S3, RNPC3, MAGEL2, which has enabled us to analyse mutation frequency in these genes within a larger proportion of our cohort simultaneously. Thereby suggesting candidate genes which may be worth pursuing functionally and may have essential roles in HP development. We have so far processed 144 patients on the targeted gene panel, and data have been aligned using 3 different pipelines; haplotype, strelka and manta. To date, ~50% of these patients have had their results analysed through variant calling, with 148 variants identified in 70/144 patients so far. These variants have been carefully filtered as potentially pathogenic mutations, contributory or causative of the CH in the patient, and have acceptable frequency rates on control databases including the GnomAD browser. Variants that do not meet our stringent criteria thresholds have been excluded. Variants of unknown significance will be functionally tested and genotype-phenotype correlations will be carefully analysed for each patient. Preliminary data suggest an oligogenic basis to many of these cases, with the involvement of multiple genes in the pathogenicity of the disease.