Congenital hypothyroidism (CH) is the most common neonatal metabolic disorder. It has an incidence ranging from 1:1,400 to 1:2,800 live births in many countries (Wassner and Brown, 2015), and it results in severe neurodevelopmental impairment if not treated early and effectively. Primary CH is usually classified into two categories by pathogenesis: thyroid dysgenesis, a defect in thyroid gland development in which a few cases were caused by FOXE1, NKX2-1, NKX2-5, and PAX8, and thyroid dyshormonogenesis (DH), an intrinsic defect of thyroid hormone biosynthesis caused by DUOX2, DUOXA2, IYD (DEHAL1), TG, TPO, SLC26A4 (PDS), SLC26A7, SLC5A5 (NIS), and TSHR (Cangul et al., 2018; Kwak, 2018). Based on the newborn screening (NBS) program and clinical diagnosis, thyroid dyshormonogenes dominate compared to thyroid dysgenesis in the Chinese population upon increased molecular diagnosis (Long et al., 2018; Sun et al., 2018). Whereas thyroid dysgenesis still accounts for more than 69% of primary CH worldwide (Wassner and Brown, 2015; Peters et al., 2018). The inheritance of CH is controversial. Although it has been understood to be autosomal recessive (biallelic) in most cases as a monogenic disorder, a few CH cases appear to be monoallelic in one gene (Nicholas et al., 2016, Fugazzola et al., 2003), or 2 or more variants in different genes (Sriphrapradang et al., 2011; Satoh et al., 2015; Makretskaya et al., 2018; Yamaguchi et al., 2020). Here, we report 10 permanent congenital hypothyroidism (PCH) cases carrying digenic variants in which each affected individual is heterozygous at two of pathogenic genes simultaneously as well as the identification of seven novel genetic variants.

During January 2015 and December 2019, the mean incidence of CH was 1:1,093 based on the NBS program in the Children's Hospital, Zhejiang University. CH screening strategies are designed to detect elevated levels of TSH and/or decreased concentrations of thyrocine (T4) (Group for Newborn Screening Society of Child Health Chinese Preventive Medicine Association, 2011). Total 2647 CH cases were diagnosed, of which 148 cases were offered genetic tests, and 66 cases (44.6%) had clear genetic confirmation, either carrying one P/LP variant in a dominate gene or two P/LP variants in a recessive gene. However, another 10 CH patients carrying digenic variants were retrospectively reviewed. They were clinically diagnosed to be PCH with a defect of thyroid hormone biosynthesis based on careful evaluation of clinical features and levothyroxine treatment during follow-up. As shown in Table 1, all patients had initially elevated TSH levels (≥9 μIU/mL), ranging from 9.15 to 25.5 μIU/mL, and were proven to be permanent by receiving a trail off levothyroxine (LT4) at 2–3 years of age. Additionally, the influences of preterm, low-birthweight, and autoimmune thyroid disease on these cases were excluded. The detailed clinical information of the patients was listed in Table 1. With LT4 treatment with a dose of 12.5–33.3 μg per day, all patients had normal ASQ (Ages & Stages Questionnaires) and maintained serum TSH levels ranging from 1 to 10 (mIU/L) with a normal level of free thyroxine (FT4) between 9.01 and 19.05 (pmol/L) cutoff during follow-up. Cases #5 has a goiter by ultrasound during NBS with dimensions 2.3 × 0.9 × 0.8 cm (Right) and 2.2 × 1.0 × 0.8 cm (Left) as previously reported (Wang et al., 2014). There was no compensatory goiter recorded in Case #5 after 1 year with the LT4 supplement.

Most cases of CH are common endocrine disorders caused by biallelic or monoallelic variants in one gene. With the widespread use of newborn screening programs and the application of genetic testing, some cases were found to carry two or more variants at different genes (Satoh et al., 2015; Nicholas et al., 2016; Sun et al., 2018; Yamaguchi et al., 2020), indicating the complexity of genetic etiology in CH. Cases with two or more variants in different genes were usually understood to be oligogenic cases, compared to those in biallelic and monoallelic cases (Yamaguchi et al., 2020). Here, we present 10 PCH cases carrying digenic variants in genes involved in thyroid hormone biosynthesis. Similar to our findings, another 58 cases harboring digenic variants were reported elsewhere (Supplementary Table 1). A total of 24 cases had digenic variants in TSHR and DUOX2, including 5 cases out of 220 Chinese CH (Fang et al., 2019) and 6 cases in Japanese patients (Abe et al., 2018; Yamaguchi et al., 2020). The coexistence of heterozygous variants in TSHR and DUOX2 was also revealed in Caucasian cases (Makretskaya et al., 2018; Sasivari et al., 2019). More digenic variants were heterozygous in two causative genes, including combined DUOX2 and TG in 13 patients (Löf et al., 2016; Fan et al., 2017; Long et al., 2018; Sun et al., 2018; Yamaguchi et al., 2020), DUOX2 and DUOXA2 in 4 patients (Zheng et al., 2016; Yamaguchi et al., 2020), DUOX2 and TPO in 3 patients (Matsuo et al., 2016; Long et al., 2018; Makretskaya et al., 2018), TG and TPO in 6 patients (Nicholas et al., 2016; Makretskaya et al., 2018; Yamaguchi et al., 2020), and TG and SLC26A4 in 2 patients (Löf et al., 2016; Sun et al., 2018). Moreover, it was also demonstrated that 23% of Italian CH patients harbored pathogenic variants in more than one gene (Filippis et al., 2017), indicating that there was no ethnicity limiting the digenic form but rather a frequency of dyshormonogenesis-associated variants. As shown in Supplementary Table 1, eight genes (TSHR, TG, DUOX2, DUOX1, DUOXA2, TPO, IYD, and SLC26A4) were present in those oligogenic cases. The higher frequency genes were DUOX2 (35.3%), TSHR (22.8%), and TG (22.8%). This was consistent with prior studies showing that DUOX2 and TSHR variants were more prevalent in Chinese, Japanese, and Korean patients (Jin et al., 2014; Fu et al., 2016; Fang et al., 2019; Yamaguchi et al., 2020). Higher frequent TG variants were detected in the Sudanese population (Bruellman et al., 2020). However, only two cases harbored variants of IYD that one individual combined with TG (Makretskaya et al., 2018) and the other one with DUOX1(Sun et al., 2018). The digenic variants thereby seemed to be common in CH, but it is somewhat challenged in the variant interpretation by the dominant effect of some of these variants. For example, there were monoallelic variants reported in DUOX2 (Moreno et al., 2002), and later this turned out to be associated with transient hypothyroidism (Wang et al., 2014; Matsuo et al., 2016).

To date, all reported genes with digenic variants are involved in the same metabolic pathway: thyroid hormone biosynthesis. As shown in Figure 2, the thyroid hormone is synthesized at the apical surface of polarized thyroid follicular cells, where the initial step is the binding of TSH to its receptors (TSHR) in the basolateral membrane, activating TG expression. To date, only 5 oligogenic cases carried heterozygous TSHR and TG variants in the Chinese and Japanese population (Fu et al., 2016; Yamaguchi et al., 2020). However, most cases of heterozygous TSHR (77.4%, 24/31) were combined with heterozygous DUOX2 as shown in Supplementary Table 1 (involved in steps 1 and 2). Subsequently, TG, TPO, and the DUOXs (DUOX2 and DUOX1) and their accessory protein DUOXA2 are involved in iodide oxidation to form T4 and T3 (Kwak, 2018). Here, 7 out of 10 cases were caused by two of the five genes in this step. Additionally, overall 32 oligogenic cases (47%, 32/68) carried the combination of two heterozygous variants in two genes in this step, indicating that iodide organification defects may be more common in CH patients. Genes, IYD/DEHAL1, NIS/SLC5A5 and PDS/SLC26A4 involved in recycling of T4, T3, iodide and tyrosine (Spitzweg et al., 2000). Limited by the number of cases, there was no variant detected in NIS as elsewhere (Long et al., 2018). Only a few variants have been reported to date in IYD and PDS. As shown in Supplementary Table 1, there were only three cases carrying heterozygous PDS and two cases carrying heterozygous IYD. Theoretically, any defects of these eight proteins in substrates, enzymes, and transport molecules in the same metabolic pathway led to thyroid dyshormonogenesis. In fact, our data and recent evidence revealed that the combinations of two pathogenic genes predominantly happened in iodide organification and then in TG expression.

The limitation here is the lack of a parental phenotype, and all our cases are simplex cases. Especially in Case#2 and #9, the parental phenotype is critical in understanding the functional effects of digenic variants. Unfortunately, the mothers carrying the same two variants refused to test their TSH levels. Moreover, the digenic variants in Case#1, #8, and #10 were classified to be VUS according to the ACMG guidelines, which need more cases or further functional experiments to evaluate their damage prediction.

Summarily, we reported here 10 PCH cases with digenic variants involved in the same metabolic pathway: thyroid hormone biosynthesis. To date, 68 CH patients have been reported harboring digenic variants in this metabolic pathway, including genes TSHR, TG, DUOX2, DUOX1, DUOXA2, TPO, IYD, and SLC26A4 with a high frequency of DUOX2, TSHR, and TG. The data present here will extend our awareness of the complexity of genetic etiology in CH.

The original contributions presented in the study are publicly available in NCBI using accession number PRJNA734721.

The studies involving human participants were reviewed and approved by the Research Ethics Committees, Children's Hospital of Zhejiang University School of Medicine. Written informed consent to participate in this study was provided by the participants' legal guardian/next of kin. Written informed consent was obtained from the individual(s), and minor(s)' legal guardian/next of kin, for the publication of any potentially identifiable images or data included in this article.

PJ and QS performed the conception, analysis, and interpretation of data. RY and LZ recruited patients and performed clinical evaluation. YL, CY, XW, and JF conducted the mutational sequencing and data analysis. PJ, YL, RY, and QS drafted and revised the article. All authors contributed to the article and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.