Cushing disease due to a somatic USP8 mutation in a patient with evolving pituitary hormone deficiencies due to a germline GH1 splicing variant

SUMMARY We present the unique case of an adult Brazilian woman with severe short stature due to growth hormone deficiency with a heterozygous G to T substitution in the donor splice site of intron 3 of the growth hormone 1 (GH1) gene (c.291+1G>T). In this autosomal dominant form of growth hormone deficiency (type II), exon 3 skipping results in expression of the 17.5 kDa isoform of growth hormone, which has a dominant negative effect over the bioactive isoform, is retained in the endoplasmic reticulum, disrupts the Golgi apparatus, and impairs the secretion of other pituitary hormones in addition to growth hormone deficiency. This mechanism led to the progression of central hypothyroidism in the same patient. After 5 years of growth and thyroid hormone replacement, at the age of 33, laboratory evaluation for increased weight gain revealed high serum and urine cortisol concentrations, which could not be suppressed with dexamethasone. Magnetic resonance imaging of the sella turcica detected a pituitary macroadenoma, which was surgically removed. Histological examination confirmed an adrenocorticotropic hormone (ACTH)-secreting pituitary macroadenoma. A ubiquitin-specific peptidase 8 (USP8) somatic pathogenic variant (c.2159C>G/p.Pro720Arg) was found in the tumor. In conclusion, we report progression of isolated growth hormone deficiency due to a germline GH1 variant to combined pituitary hormone deficiency followed by hypercortisolism due to an ACTH-secreting macroadenoma with a somatic variant in USP8 in the same patient. Genetic studies allowed etiologic diagnosis and prognosis of this unique case.

SELAdb: A database of exonic variants in a Brazilian population referred to a quaternary medical center in São Paulo

OBJECTIVES: High-throughput sequencing of genomes, exomes, and disease-focused gene panels is becoming increasingly common for molecular diagnostics. However, identifying a single clinically relevant pathogenic variant among thousands of genetic polymorphisms is a challenging task. Publicly available genomic databases are useful resources to filter out common genetic variants present in the population and enable the identification of each disease-causing variant. Based on our experience applying these technologies at Hospital das Clínicas da Faculdade de Medicina da Universidade de São Paulo (HCFMUSP), São Paulo, Brazil, we recognized that the Brazilian population is not adequately represented in widely available genomic databases. METHODS: Here, we took advantage of our 5-year experience as a high-throughput sequencing core facility focused on individuals with putative genetic disorders to build a genomic database that may serve as a more accurate reference for our patient population: SELAdb. RESULTS/CONCLUSIONS: Currently, our database comprises a final cohort of 523 unrelated individuals, including patients or family members managed by different clinics of HCFMUSP. We compared SELAdb with other publicly available genomic databases and demonstrated that this population is very heterogeneous, largely resembling Latin American individuals of mixed origin, rather than individuals of pure European ancestry. Interestingly, exclusively through SELAdb, we identified a spectrum of known and potentially novel pathogenic variants in genes associated with highly penetrant Mendelian disorders, illustrating that pathogenic variants circulating in the Brazilian population that is treated in our clinics are underrepresented in other population databases. SELAdb is freely available for public consultation at: http://intranet.fm.usp.br/sela