Molecular and Clinical Features of Congenital Hypothyroidism Due to Multiple DUOX2 Variants

Background: DUOX2 is one of the major causative genes of congenital hypothyroidism (CH). Still, the mutation spectrum and clinical outcomes of biallelic DUOX2 variants are not fully understood. This study aimed to elucidate the molecular features and long-term clinical manifestations of CH caused by multiple pathogenic DUOX2 variants. Methods: A total of 255 patients with CH were screened for rare variants of 11 known causative genes. DUOX2 variants were classified according to their protein structure and residual activity. In vitro assays were performed for several variants of unknown functions. Clinical analyses were conducted for patients with multiple pathogenic variants of DUOX2 but not of other genes. Results: We identified 24 pathogenic variants of DUOX2, together with two benign variants and seven variants of uncertain significance, in 63 patients. The pathogenic variants included three missense substitutions and one frameshift variant that have not yet been linked to CH. Twenty-one patients carried multiple pathogenic DUOX2 variants without any other pathogenic gene variants. Three of the 21 patients harbored homozygous variants. Family analysis, long-read amplicon sequencing, and haplotype phasing confirmed compound heterozygosity of the DUOX2 variants in 14 patients, whereas the allelic positions of the variants in the remaining four patients could not be determined. Of the 21 patients, 19 were treated with levothyroxine; their ages at drug withdrawal ranged from 9 months to 21.4 years. Three patients required retreatment after drug-free intervals of 6 months, 8 months, and 10 years. There were no differences in clinical severity among patients with DUOX2 amorphic/amorphic, amorphic/hypomorphic, and hypomorphic/hypomorphic variants. Conclusions: These results broaden the mutational spectrum of DUOX2. Furthermore, our data imply that patients with multiple pathogenic DUOX2 variants typically exhibit transient CH without significant genotype-phenotype correlations. Most importantly, this study demonstrated for the first time that these patients are at risk of developing recurrent hypothyroidism after a long drug-free interval.