Congenital hypothyroidism (CH) is the most common endocrine disease of newborns. Its incidence varies between 1:2000 and 1:4000 in the world (1). Many countries conduct newborn screening (NBS) tests for CH, resulting in a rapid diagnosis and appropriate treatment for numerous patients, thus preventing neurodevelopmental, growth, and behavioral problems associated with untreated CH. The nationwide NBS program of CH, implemented in 1990 in Argentina, identifies children with high TSH levels, with an incidence of 1:2,367 (2).

The incidence of CH has doubled over the last two decades, particularly in cases with gland-in-situ (GIS). Several factors have been proposed to explain this upward trend, including the lowering of the thyrotropin (TSH) cut-off value for CH, the increased survival of extremely preterm infants, who are at greater risk for CH, and the higher prevalence of consanguineous families in several countries (3, 4).

The etiology of primary CH has been traditionally classified as thyroid dysgenesis (TD) or thyroid dyshormonogenesis (TDH). TDH occurs due to any defect in the steps of hormone biosynthesis and usually presents with a structurally normal-sized or goitrous thyroid. TDH constitutes 10–15% of CH patients and is often familial. TD usually occurs because of an ectopic thyroid or athyreosis, with thyroid hypoplasia as the least common cause (5, 6).

Subclinical hypothyroidism (SCH) or Mild Isolated Hyperthyrotropinemia (MIH) is usually defined as a serum TSH concentration above the biochemically defined upper limit of the reference range when serum-free T4 (FT4) concentration is within its reference range and characterized by a normal-sized or hypoplastic thyroid gland (7, 8). Of note, in the literature, the definition of SCH, and even that of CH, can be arbitrary and vary between studies.

Newborns with MIH may have transient impairment of thyroid function due to environmental or mother-derived causes. During childhood, the main cause of persistent MIH is autoimmune disease. Iodine deficiency, obesity, non-thyroidal chronic diseases, or inherited syndromes may also be responsible for mild increases in TSH levels. Persistent MIH may be the expression of mild persistent thyroid failure due to genetic abnormalities of thyroid structure or function.

The studies on the genetics of CH have increased with the advent of Next-Generation Sequencing (NGS), which allows simultaneous analysis of known and candidate genes associated with various diseases. Recent NGS studies have been useful in elucidating the genetic causes of cases of CH (9–13) with a diagnostic yield ranging from 20 to 65% (13–15). Current consensus guidelines on CH recommend that genetic testing should aim to improve the patient’s diagnosis, treatment, or prognosis (16).

DUOX2 (OMIM: 606759), TG (OMIM: 188450), and TPO (OMIM: 606765) are the genes most frequently implicated in cases of TDH; however, other genes have also been included in the “Congenital Hypothyroidism: A 2020–2021 Consensus Guidelines Update” as genes associated with TDH, such as DUOX1, DUOXA2, SLC5A5 (NIS), SLC26A4 (PDS), SLC26A7, IYD (DEHAL1), GNAS, and TSHR (OMIM: 603372) (16). DUOX2 variants appear to be the most common cause of permanent TDH in East Asia (13, 17–19). Instead, studies on the Caucasian population showed that TPO or TG variants are the most common cause of TDH (10, 14). Biallelic variants in TG and TPO may result in goiter and mild to severe CH (11, 20, 21). Homozygous or compound heterozygous variants in the TPO gene lead to a variable degree of iodine organification deficiency characterized by hypothyroidism with elevated thyroglobulin levels (22).

Of note, the TSHR gene has a special place in the etiology of primary CH and is involved in TD, TDH, and Resistance to TSH. This last condition of reduced sensitivity to TSH is characterized by elevated serum TSH concentration in the absence of goiter, with a normal or hypoplastic thyroid gland, and normal to very low levels of thyroid hormones (23). Patients with TSHR variants present a phenotype varying from persistent MIH with normal-sized glands to severe TD with orthotopic hypoplasia (14, 24–26).

Isolated thyroid hypoplasia may occur because of variants in PAX8 or TSHR ( 5 ). Although thyroid morphology and degree of biochemical hypothyroidism are very variable in the context of PAX8 variants, patients have been described with MIH or isolated hypoplasia (27). Variants in the transcription factors NKX2–1 and FOXE1 cause CH in association with more extensive developmental syndromes (5).

With the widespread use of NBS programs and the application of NGS testing, it has become evident that the genetic architecture of CH is more complex than previously anticipated. A few CH cases appear to be monoallelic, associated with milder phenotypes (10, 11). For DUOX2 and TSHR authors have applied both autosomal recessive and autosomal dominant inheritance (10). Data on oligogenic combinations with one or more variants in different genes in CH have started to accumulate in scientific literature (28).

This study aimed to explore the underlying genetic etiology of MIH in a well-characterized pilot prospective cohort, using a 13-gene virtual NGS-based panel for variant detection. Additionally, to increase the knowledge on the MIH variation landscape and provide insights into the pathogenesis of the disease, we performed a systematic review of variants reported as causal for MIH in scientific literature up to 8 January 2025. Variants were systematically collected, and pathogenicity was reassessed.

Molecular diagnosis was highly suggested in 4/10 patients, indicating a modest diagnostic performance for MIH with a 13-gene virtual NGS panel.

The proportion of patients with GIS who receive a molecular diagnosis varies widely across studies. Factors contributing to this variability include differences in patient phenotypes, clinical characterization of the patients, and ethnicity, but especially differences in variant classification and interpretation of the genetic test. Although the ACMG/AMP guidelines established a widely adopted five-tier system and a framework of evidence-based criteria for classifying variants in clinical practice (32), a consensus-structured standard that ensures evidence-based classifications is currently missing for most gene-disease or gene-phenotype associations. Our work represents an important, initial step in further facilitating the interpretation of genomic data from MIH. We proposed a variant classification framework and genetic diagnostic criteria for MIH. Furthermore, we provided the research community with a unique and accurately annotated database of MIH variants, mapped to functionally relevant transcripts in GRch37 to facilitate future investigations.

Different studies have reported monoallelic variants in TDH using NGS technology (26). The prevailing understanding is that most cases of TDH are inherited in an autosomal recessive manner; however, hypotheses have been raised about the heritability of CH, especially CH associated with variants in TSHR and DUOX2 genes, which appear to follow autosomal recessive and autosomal dominant patterns (10). Isabelle Oliver-Petit and col. have recently reported heterozygous variants in the TG, TPO, DUOX2, and TSHR genes associated with moderate and mild CH (11).

In several reports, TSHR variants were regarded as an autosomal dominant trait (38, 39). Almost two decades ago, it was suggested for the first time that a mechanism of negative dominance by direct interaction of the TSHR mutant with wild-type receptors as an explanation for the dominant inheritance of partial TSH resistance (38). Different authors have reported that patients with monoallelic TSHR variants exhibited a heterogeneous clinical presentation ranging from subclinical to severe CH (11, 40). Patient 9 is illustrative of a chronically slightly elevated TSH value associated with a monoallelic TSHR variant.

TG or TPO deficiency is generally inherited in an autosomal recessive manner, and affected patients have either homozygous or compound heterozygous variants, so thyroid dysfunction is not expected in heterozygous individuals. However, there are some reports of TDH caused by monoallelic variants in TPO or TG (11, 41). Fugazzola L (2003) reported a case of three siblings with severe CH due to a total iodine organification defect caused by a monoallelic TPO variant (42). Nicholas AK (2016) has reported four cases of subclinical or mild CH, classified as unsolved or ambiguous, harboring heterozygous sequence variants in TG or TPO (10). One study recently reported monogenic heterozygous TG variants in three patients. The authors demonstrated that the pathogenic variation co-segregated with the phenotype in one family, suggesting it is possibly causative for CH (14).

We presented here patient 5 with a monoallelic novel variant in the TG gene. Excluding this important finding, which has also expanded the TG variation spectrum, a single rare variant is not sufficient to explain the pathophysiology of MIH, and the case remains unsolved.

Our systematic review showed that monoallelic clinically significant variants in TG or TPO were rarely reported (N = 8, Table 4 ). Conversely a ~70% of monoallelic variants of clinical relevance in the TSHR and DUOX2 genes reported to be associated with MIH may explain the disease genetically.

In a model of complete penetrance and expressivity, the pathogenetic variant should co-segregate with the phenotype in the family; however, the diagnosis of MIH is determined mainly by laboratory evaluations, and most patients exhibit few or no signs or symptoms of thyroid dysfunction. The observation that some maternal DUOX2 mutation-carriers are euthyroid in adulthood supports the transiency of CH in most affected cases (19). Higher TSH screening cut points 20 years ago may have failed to diagnose borderline CH in carrier parents.

In our systematic review of the literature, we collected information about the family history of thyroid disease and segregation analyses when available. Co-segregation with phenotype was verified in 5 compound heterozygous patients. On the other hand, 34 monoallelic variants were also detected in affected parents or siblings, suggesting their pathogenicity in the heterozygous state.

Monoallelic variants have previously been described in association with TDH but are usually assumed to coexist with an additional undetected CNV, intronic, or regulatory variant on the other chromosome. Technical limitations of NGS could explain the absence of a second variant in these patients (25).

Variable expressivity (clinical or phenotypic heterogeneity) in cases harboring similar causative variants suggests that mono- and oligogenic factors, as well as environmental modulators, may play a role in determining disease severity (10). Endocrine disruptors and environmental factors (e.g., iodine intake, acquired thyroid disorders, and drugs affecting thyroid function) could contribute to endocrine dysfunction by influencing gene expression and generating a more profound phenotype in carriers of rare genetic variants (9). Epigenetic factors (11) and autosomal monoallelic expression could not be excluded.

Recent NGS studies propose that digenic or oligogenic effects may play an important role in the pathogenesis of TDH (10, 14, 17). (9, 11, 13, 18) Similarly, oligogenicity is proposed for other endocrine diseases such as Congenital Hypogonadotropic Hypogonadism (43).

In contrast to the monogenic inheritance hypothesis, the true digenic inheritance refers to the phenomenon where the phenotypic expression of a disorder is influenced by disease-causing variants in two different genes. In true digenic inheritance, the simultaneous presence of variants in two genes is necessary to manifest a particular phenotype or disease (44). To account for variable expressivity, some authors have invoked the concept of genetic modifiers, which are considered, in general, to operate as additional molecular defects that determine the final phenotype on the background of the primary pathogenic variant. It has been suggested that multiple partial enzyme deficiencies may lead to clinically relevant biochemical derangements - synergistic heterozygosity - in CH (11). Aycan Z and col. have described for the first time cases with digenic DUOX1/DUOX2 variants causing complete DUOX isoenzyme deficiency in the context of likely iodine deficiency. These individuals manifest severe CH, suggesting failure to compensate for defective thyroid H2O2 synthesis (45). Yang R and col. have recently reported 10 cases carrying digenic variants in genes involved in TDH, along with a literature review on digenic variants recording 58 cases with CH (28). One hundred and eighteen combinations are linked to CH in the OLIgogenic diseases DAtabase (OLIDA, https://olida.ibsquare.be, last accessed September 2025), a curated database of oligogenic diseases and their gene variants.

In this study, presumed digenic or pseudo-digenic inheritance was detected in three cases from the pilot cohort study. A definite diagnosis was established in patient 6 based on our genetic diagnosis criteria ( Supplementary Table S4 ) when the monoallelic variant classified as LP in the DUOX2 gene is considered (NM_001363711.2:c.2895_2898del). However, assuming a monogenic plus modifier scenario, the P variant in the TG gene can be regarded as a putative genetic modifier (NM_003235.5:c.886C>T). The primary disease-causing variant and the modifier together may better explain the clinical picture of the patient than each variant alone. Similarly, the phenotype in patients 2 and 8 is explained by a heterozygous variant in the TSHR gene in combination with a modifier in the TPO gene. Through systematic review we have identified three cases of likely oligogenic inheritance associated with MIH ( Supplementary Table S6.6 ). Interestingly, in two individuals, monoallelic variants in the DUOX2 or TSHR genes can be reasoned as the primary disease-causing variants in a presumed pseudo-digenic scenario. In the third patient, a deleterious variant combination was reported in the TPO and TG genes, but probably more evidence is needed to establish a true digenic model.

Further research is needed to clarify the role of rare TG and TPO variants in digenic inheritance or if they may act as disease modifiers of the phenotype. In general, more studies with large pedigrees and clear phenotypic variability investigating oligogenic involvement in CH are required (25, 40).

Digenic variants appeared to be common in CH, but they challenge variant interpretation and make clinical diagnosis difficult. The analysis of oligogenic combinations is currently not subjected to any standards and guidelines and requires different considerations from the framework developed for the interpretation and reporting of variants implicated in Mendelian diseases.

The treatment of CH is not conditional on identifying genetic etiology. However, personalized treatment strategies should consider the individual genetic background, potential compensatory mechanisms, and overall, the clinical context, to optimize care for all patients in the broad spectrum of thyroid hypofunction. A recent study reported that genetic testing influenced treatment decisions for patients with permanent CH (14).

Uncertainty about the benefits of levothyroxine therapy still exists for partial resistance to TSH and compensated euthyroid MIH in children (7, 46). L-T4 treatment seems reasonable in patients with childhood SCH who have TSH levels >10 µU/ml during the first 3 years of life when most thyroxine-dependent brain maturation has occurred (46). Then, a follow-up plan should be tailored to the specific circumstances of the child. Leonardi et al. have stated that when serum TSH is higher than normal during early childhood, a high risk (~50%) of persistent SCH is present in children with genetic and/or morphological abnormalities and should be reassessed periodically (47). Previous studies have shown that untreated children with well-compensated hyperthyrotropinemia due to monoallelic TSHR genetic alterations have normal growth and development (48–50). Tenenbaum-Rakover et al. studied the long-term outcome (over 11 years) of loss-of-function variants in the TSHR gene and showed that SCH in heterozygous subjects is a stable compensated condition with an appropriately adjusted pituitary TSH set point and does not require replacement therapy (51). However, homozygous and compound heterozygous subjects may necessitate L-T4 therapy because of incompletely compensated SCH (49, 51).

Genetic information allows the individualized decision to follow up without treatment for mildly elevated TSH in childhood. For patients with a definitive or inconclusive molecular diagnosis, such as our case study, periodic biochemical surveillance through puberty, young adulthood, and pregnancy should be implemented to ensure early identification of patients who might benefit from treatment. Patients and mutation-carrier family members should be counseled regarding the possible future risk of overt hypothyroidism, particularly during pregnancy, when iodine status can also be compromised. Furthermore, patients with MIH could evolve into clinical hypothyroidism in case of acquired thyroid disorders, such as the onset of autoimmunity (46). Finally, familial genetic counseling may also help in recognizing the threat of CH in the event of compound heterozygous carriers. There are some reports of cases carrying a heterozygous TSHR variant associated with adult-onset compensated TSH resistance having compound heterozygote children with CH (52).

This study has some limitations that should be considered. The main restriction of our study is the limited number of patients. Yet, as our hospital operates as a referral medical facility specializing in the treatment of CH, the results still hold a certain degree of representation. Samples were obtained from parents for segregation analysis; however, co-segregation of the variant with the disease phenotype could not be verified in all cases. Iodine status was not assessed. The requirement for ongoing levothyroxine replacement (case 7) or continuing TSH elevation (cases 1 and 3) suggested persistent MIH in at least three unsolved cases. Two other unsolved cases have not yet undergone a formal trial-off levothyroxine withdrawal (patients 4 and 10). It is also conceivable that despite adequate median coverage, nonuniform coverage of genes could have failed to detect variants (type II error). It was verified that, in general, the coverage of the exons of interest at 20x is at least 99%, and uncovered regions still have a coverage of 10x >90%. No functional analysis was carried out in this study to evaluate the mechanism by which monoallelic variants resulted in functional impairment. Finally, beyond inherent limitations of the systematic review process, the major constraint we encountered was the differing sequencing methodologies, with a relatively low number of cases investigated with the NGS approach.

Our study provides new insights into the genetic etiology of MIH and expands the phenotypic and variant spectrum of CH. The pilot cohort demonstrated a modest diagnostic yield using a 13-gene NGS panel, while the systematic review offers a comprehensive resource of clinically relevant variants, with TSHR, DUOX2, TPO, and TG being the most frequently affected genes. The overall results suggest that the MIH phenotype is influenced not only by monoallelic and biallelic variants but also potentially by digenic variant combinations. Further studies are required to delineate the clinical significance of monoallelic variants in MIH, especially in TG and TPO genes.

Overall, these findings highlight the importance of comprehensive genetic testing for personalized follow-up and management and emphasize the need for larger multicentric prospective studies to investigate the utility of NGS in guiding treatment decisions and predicting prognosis for MIH patients.