Approach to the patient: genetics and management of congenital hypothyroidism

Abstract

Congenital hypothyroidism (CH) encompasses a diverse spectrum of disorders with diverse genetic etiologies and variable clinical courses, ranging from transient neonatal hyperthyrotropinemia to permanent thyroid hormone or thyrotropin deficiency. Advances in molecular genetics have substantially expanded the catalog of genes implicated in CH; however, the translation of this knowledge into everyday clinical practice remains challenging. In this Approach to the Patient, we use selected clinical vignettes to illustrate how genetic information can substantiate the diagnosis, management, and counseling of patients with CH. Rather than providing an exhaustive genetic review, this paper focuses on clinically relevant scenarios and pragmatic decision points—when to perform genetic testing, how results influence treatment duration and intensity, and how they guide prognostication and family counseling. This case-based framework emphasizes a pragmatic, patient-centered approach to the use of genetics in the management of CH.

Background

Thyroid hormones (TH) are essential for normal growth, metabolism, and neurodevelopment. An early as possible diagnosis of congenital hypothyroidism (CH) and prompt levothyroxine treatment is therefore a major public health priority. Since 1974, neonatal screening, generally by measuring TSH, has enabled early detection of primary CH, although central CH (CeCH) may be missed (1). CH affects about 1/2500-3000 newborns worldwide (2) and is more common in girls than in boys (2:1 ratio for thyroid developmental defects) and in Caucasians than in Black people (3).

TH synthesis and secretion are tightly regulated by the hypothalamic (TRH)-pituitary (TSH)-thyroid (TH) axis. Thyroid development begins around the third gestational week, with the emergence of a median endodermal anlage, followed by migration, fusion with the ultimobranchial bodies, and differentiation into hormonally competent follicular cells. The specification process is orchestrated by transcription factors including NKX2-1, PAX8, FOXE1, and HHEX (4, 5). Terminal differentiation is characterized by the expression of key components of the TH biosynthesis pathway, such as the TSH receptor, sodium-iodide symporter (NIS/SLC5A5), thyroglobulin (TG), thyroperoxidase (TPO), SLC26A4, IYD, and DUOX2/DUOXA2 complex. Disruption at any developmental or functional stage, ranging from specification and migration to hormone synthesis, can result in CH (6).

Syndromic primary congenital hypothyroidism

CH may occur as part of a syndromic genetic disorder in which thyroid dysfunction results from impaired signaling or from primary defects of thyroid development or hormone biosynthesis.

The most common cause of primary CH is thyroid dysgenesis (TD), with 60% of cases (2). TD categories are thyroid ectopy, due to an embryonic migration defect; thyroid agenesis, due to disappearance or abnormal differentiation of thyroid progenitor-cells during embryonic life; thyroid hypoplasia and hemi-agenesis. In 95% of cases, no molecular defect can be identified. Few genes related to TD have been identified (NKX2-1 (7), PAX8 (8), FOXE1 (9), TSHR (10), GLIS3 (11), NTN1 (12), JAG1 (13), CDCA8 (14), TUBB1 (15), and more recently TRPC4AP (16), GBP1 (17), and EIF4B (18)).

Moreover, the risk of CH is increased in several genetic syndromes with predominantly extrathyroidal features, including Down syndrome/Trisomy 21 (DYRK1A), pseudohypoparathyroidism (GNAS), DiGeorge syndrome (TBX1), Ohdo syndrome, genitopatellar syndrome (KAT6B), Townes-Brocks syndrome (SALL1), Kabuki syndrome (KMT2D/MLL2, KDM6A), Johanson–Blizzard syndrome (UBR1), Williams-Beuren syndrome (7q11.23 microdeletions), Takenouchi-Kosaki syndrome (CDC42), and blepharo-cheilo-odontic syndrome (CDH1/CTNND1) (6). The mechanisms underlying CH in these syndromes remain unclear, and the genes involved are typically ubiquitously expressed, including within the thyroid gland. Figure 1 illustrates the most common extrathyroidal features associated with genes causing syndromic primary congenital hypothyroidism.

Figure 1.

Most common extrathyroidal features associated with genes causing syndromic primary congenital hypothyroidism. Created in BioRender. NGUYEN QUOC, A. (2026).

Recent advances in next-generation sequencing (NGS), combined with detailed phenotypic characterization of patients and families, have considerably improved our understanding of CH genetics. Here, we focus on the clinical integration of genetics into the diagnosis and management of CH. Several recent reviews extensively discuss the molecular pathways and gene catalogs associated with CH (6, 19). Therefore, we will focus on when and how genetic information can guide clinical decision-making. We describe six patients. All five patients and their legal representatives/parents have consented to having their anonymized cases reported.

Clinical case reports

Patient #1

Case report

A female newborn was referred to our tertiary neonatal intensive care unit because of severe respiratory distress immediately after birth, requiring endotracheal intubation within the first two hours. She was the first child of non-consanguineous parents. The pregnancy had been uneventful. She was born at term with birth weight and length appropriate for gestational age.

Routine neonatal screening on day 3 of life showed high TSH level (78 mIU/L; normal, ≤17 mIU/L). On day 5, TSH elevation persisted (81 mIU/L; normal, 0.3-7 mIU/L) and free thyroxine (fT4) was borderline low (10.5 pmol/L; normal, 11-22 pmol/L), consistent with primary CH. Levothyroxine therapy, 10 µg/kg/day, was initiated on day 5. Thyroid ultrasonography on day 11 showed a normal-sized eutopic thyroid gland. Thyroid autoantibodies were negative, thyroglobulin was detectable, and iodine status investigations did not suggest iodine deficiency or overload. The mother had normal thyroid-function tests and no thyroid autoantibodies. TH replacement ensured euthyroidism. At one month of age, however, the girl remained hypoxemic and hypotonic, with insufficient weight gain. High-resolution chest computed tomography demonstrated diffuse ground-glass opacities. Marked neutrophilic alveolitis without evidence of infection was diagnosed by bronchoscopy with bronchoalveolar lavage, and a lung biopsy showed severe alveolar involvement with septal and luminal inflammation. Genetic testing was negative for surfactant dysfunction disorders (SFTPB, SFTPC), cystic fibrosis, and other common causes of neonatal interstitial lung disease.

Chronic respiratory failure with persistent oxygen dependency developed. When she was 18 months of age, an intercurrent infection caused a fatal episode of acute respiratory failure.

Further discussion of patient #1

The presence of primary CH with a eutopic thyroid gland and severe neonatal pulmonary disease prompted targeted genetic testing at 6 months of age, which identified a heterozygous, de novo, missense mutation in the NKX2-1 gene (NM_003317.4 c.493C>T, p.Arg165Trp), indicating a diagnosis of Brain-Lung-Thyroid syndrome (20, 21).

NKX2-1 encodes a major transcription factor expressed during early thyroid, lung, and forebrain development (5). It is involved in thyroid differentiation, via the regulation of key genes (TSHR, TG, TPO, SLC5A5, SLC26A4), and in maintaining the ordered architecture and function of the differentiated thyroid (4, 22). In the lung, NKX2-1 is implicated in the branching of lobar bronchi and in regulating surfactant-protein expression in type II pneumocytes. NKX2-1 is central in the specification and migration of interneuron subtypes in the developing embryonic brain (23, 24). The full triad is Brain-Lung-Thyroid syndrome and occurs in about 50% of patients with pathogenic NKX2-1 variants. Other patients have variable combinations of thyroid, lung, and/or brain abnormalities. Neurological symptoms occur in 93-100% of patients and typically consist of benign hereditary chorea, although hypotonia, ataxia, and developmental delay may occur. Pulmonary manifestations, present in 46-78% of patients, range from infant respiratory distress syndrome to recurrent lung infections. Finally, thyroid abnormalities are noted in 67–87% of patients; the most common are mild hypothyroidism or isolated hyperthyrotropinemia with normal thyroid morphology (50% of cases), although thyroid hypoplasia, hemi-agenesis, and athyreosis have been reported (25-27). TSH-based neonatal screening may be negative, as TSH elevation may be delayed. Pathogenic NKX2-1 variants often occur de novo but can be inherited on an autosomal dominant basis. Single nucleotide variants are the most common abnormalities, although NKX2-1 deletion can occur. A report of deletions proximal to NKX2-1 suggests the presence of an upstream enhancer in this region (28). Although pathogenic variants generally lead to haploinsufficiency, a dominant negative effect has been reported with some variants (25, 29).

Genetic testing is mandatory when brain–lung–thyroid syndrome is suspected, given the frequently incomplete and misleading presentation and possible false-negative neonatal TSH screening test. Neurological manifestations, particularly benign hereditary chorea and/or pulmonary symptoms, may predominate initially. The molecular diagnosis enables anticipatory clinical management, including targeted monitoring for potentially severe or life-threatening pulmonary complications, which are the main source of morbidity and mortality (25). Lung cancer has been reported in patients with nonsense pathogenic variants (30, 31). A patient with lung cancer had no history of pulmonary symptoms, indicating a need for pulmonary monitoring over time (30). The molecular diagnosis also allows a more accurate neurological prognosis, prevents misdiagnosis as a neurodegenerative disease, and guides early developmental interventions. For the family, identifying a pathogenic variant, which is usually de novo, allows appropriate genetic counseling. Finally, because brain–lung–thyroid syndrome can be caused by NKX2-1 deletions or upstream regulatory defects, diagnostic strategies capable of detecting copy-number variations are required to avoid missing the diagnosis. In conclusion, appropriate genetic testing makes crucial contributions to the diagnosis, prognosis, and multidisciplinary care of patients with suspected brain-lung-thyroid syndrome.

Patient #2

Case report

A male infant was referred to our tertiary pediatric endocrinology unit for follow-up of CH identified neonatally. He was conceived by in vitro fertilization and had severe intrauterine growth restriction, with short femurs noted during the third trimester. Results of prenatal testing were normal, including fetal karyotype (46, XY); fluorescence in situ hybridization for trisomies 13, 18, and 21; FGFR3 sequencing; and screening for autoimmunity. He was born at full term by cesarean section due to acute fetal heart-rate abnormalities. At birth, weight was 2700 g (28th percentile), length 44 cm (1.8th percentile), and head circumference 32.5 cm (16.6th percentile). Transient hypoglycemia (nadir 2.1 mmol/L) managed with increased caloric intake, resolved by day 10.

The mother was taking levothyroxine for hypothyroidism and had a history of pancreatic neuroendocrine tumor. Her height was 150 cm (−2.3 SD according to French growth charts). The father had no significant medical history and the parents were not consanguineous.

Cryptorchidism was noted and, together with the neonatal hypoglycemia, prompted an early endocrinological evaluation for suspected hypopituitarism. On postnatal day 3, fT4 was in the low-normal range (19.5 pmol/L; normal, 11.6-36.0 pmol/L). Repeat testing on day 11 showed TSH elevation (27.2 mIU/L; normal, 0.3-7 mIU/L) and an fT4 level of 17.7 pmol/L; normal, 9.5-28.9 pmol/L), consistent with primary CH. Tests for thyroid autoantibodies were negative and serum TG was detectable.

Levothyroxine therapy, 5 µg/kg/day, was initiated on day 12. One week later, fT4 and TSH were normal (21.4 pmol/L and 4.25 mIU/L, respectively). Subsequent dosage adjustments maintained biological euthyroidism during follow-up.

Thyroid ultrasonography showed a normally-sized eutopic thyroid gland with a uniform echotexture and multiple bilateral millimetric cystic formations but no nodules. Thyroid scintigraphy, performed after levothyroxine therapy had normalized the TSH level (4.25 mIU/L), showed low iodine-123 uptake by a normally located gland.

Subcutaneous ectopic ossifications were detected at the age of 12 months. At 17 months, brachydactyly affecting several digits was noted, with shortening of the metacarpals and phalanges. Catchup growth occurred, the head circumference followed the mean, and psychomotor development was appropriate for age. Given the suspicion of hormone resistance, at age 18 months, endocrine assessment was performed and showed a normal insulin growth factor-1 (IGF-1) level (100 ng/mL; normal, 46-150 ng/mL), marked parathyroid hormone (PTH) elevation (11.43 pmol/L; normal, 1.06-5.3 pmol/L), and normal values for serum calcium and 25-OH-vitamin D.

Further discussion of patient #2

Given the intrauterine growth restriction, skeletal malformations, subcutaneous ectopic ossifications, and TSH/PTH resistance, genetic testing was performed. A pathogenic variant in the GNAS gene (NM_000516, c569_570del, p.Tyr190Cysfs*19) was identified. This variant causes a frame shift resulting in a premature codon stop. GNAS gene variants like these are responsible for Pseudohypoparathyroidism (PHP) Type 1 (32, 33). GNAS is an imprinted complex locus on chromosome 20q13.32, and PHP is caused by pathogenic variants in the allele of maternal origin. Further investigations confirmed maternal inheritance.

Pseudohypoparathyroidism, now classified within the inactivating PTH/PTHrP signaling disorders (iPPSDs), is a prototypical example of syndromic primary CH with predominant extrathyroidal manifestations (34). PHP is caused by genetic and/or epigenetic alterations of the GNAS locus, which encodes the stimulatory G protein α-subunit (Gsα), a key mediator of G protein-coupled receptor signaling involved in the TSH receptor pathway. Due to tissue-specific genomic imprinting, loss of maternal Gsα expression leads to hormone resistance in selected organs. PHP type 1A, caused by a heterozygous maternal inactivating GNAS mutation, is characterized by multihormonal resistance, typically with PTH resistance, obesity, short stature, and osteodystrophy. Mild TSH resistance is common and may be detected by neonatal screening but often develops only during childhood. PHP type 1B is due to epigenetic defects affecting differentially methylated GNAS regions and manifests predominantly as renal PTH resistance, with TSH resistance in about half the cases.

Related disorders affecting the Gsα/cAMP signaling pathway include acrodysostosis due to PRKAR1A or PDE4D mutations. The existence of these conditions further supports the central role of postreceptor signaling defects in the pathogenesis of CH (35).

Overall, iPPSDs, including PHP, indicate that CH can arise from defective intracellular TSH signaling rather than from structural thyroid abnormalities. Thus, syndromic and signaling-based etiologies must be considered in patients with CH or isolated neonatal hyperthyrotropinemia, particularly when associated with extrathyroidal abnormalities.

Patient #3

Case report

A male neonate born at 39 weeks by uncomplicated vaginal delivery was referred on postnatal day 13 for evaluation of a positive neonatal CH screening test with a palpable goiter (Fig. 2A). Both parents and a two-year-old sibling were healthy, but there was a family history of thyroid disease in the maternal great-grandparents. No data were available on maternal thyroid function during pregnancy.

Figure 2.

(A) Neonate with palpable goiter. (B) Lateral view of cervical MRI shows the goiter. (C) Bilateral deafness with bilateral enlarged vestibular aqueduct on head CT scan. Red arrows show the enlarged thyroid volume. Created in BioRender. NGUYEN QUOC, A. (2026).

At birth, the weight was 3400 g, length 51 cm, and head circumference 35 cm. Maternal cytomegalovirus (CMV) seroconversion during the first trimester of pregnancy was treated with valacyclovir until 26 weeks' gestation, when a CMV PCR test on amniotic fluid was negative. Serial prenatal sonograms were normal. At birth, the urinary CMV PCR was positive (3.76 log IU/mL). Cranial ultrasound and a fundoscopic examination were normal.

The only neonatal abnormalities were the goiter diagnosed at birth and transient jaundice requiring phototherapy. Screening on postnatal day 3 showed TSH elevation (31-34 mIU/L) and bilaterally absent otoacoustic emissions. On postnatal day 13, mild axial hypotonia, less vigorous feeding, and a palpable noncompressive goiter were noted, without respiratory or digestive symptoms. The fT4 level was 1.10 ng/dL (normal, 0.7-1.48) and fT3 was 6.1 pg/mL (normal, 1.7-3.7); the sample was insufficient for a TSH assay. Tests were negative for anti-TPO and anti-TG autoantibodies.

Close follow-up was scheduled. On postnatal day 16, muscle tone had improved and weight gain was appropriate. TSH was mildly elevated (5.34 µIU/mL), fT4 normal (1 ng/dL), and fT3 slightly elevated (5.5 pg/mL). Cervical ultrasound revealed a markedly enlarged, homogeneous thyroid gland with pronounced bilateral hypervascularity, consistent with a diffuse goiter (right lobe: 29*15*35.5 mm, volume: 8.34 mL and left lobe: 29*17*34.4 mm, volume: 8.93 mL, total volume: 17.3 mL, normal ≅1 mL). Cervical MRI confirmed the goiter (Fig. 2B). No nodules or lymphadenopathy were found. At the same time, the mother had normal thyroid function and negative tests for thyroid autoantibodies. The size of the thyroid decreased gradually and, at one month of age, the patient had normal thyroid function tests and no goiter by physical examination. Auditory brainstem responses confirmed the sensorineural hearing loss. Magnetic resonance imaging (MRI) and high-resolution computed tomography of the temporal bones suggested Pendred syndrome with an enlarged vestibular aqueduct (Fig. 2C).

At the last follow-up at two years of age, the patient had normal thyroid function, with no goiter. Bilateral cochlear implantation was performed, given the profound hearing impairment.

Patient #4

Case report

An 11-day-old female newborn of nonconsanguineous Indian parentage was referred to our department following a positive neonatal screening for primary congenital hypothyroidism (CH). Neonatal screening revealed a significantly elevated TSH of 94.18 mUI/L (reference range: <17 mUI/L). The pregnancy was complicated by diet-controlled gestational diabetes. Delivery was uneventful, and birth measurements were within normal limits. Hearing tests conducted on day 3 were normal. Upon referral, the infant exhibited axial hypotonia and a persistently open posterior fontanel. Confirmatory laboratory tests established a diagnosis of severe primary CH, characterized by high TSH level: 298 mUI/L (N: 0.8-8.5), low fT4: 3.5 pmol/L (N: 8.3-18.6) and low-normal fT3: 4.5 pmol/L (N: 4-8.5). Thyroid ultrasound revealed a goiter with a thyroid volume of 3.5 mL (N: <1 mL) but otherwise normal morphology (Fig. 3A). Thyroid Scintigraphy (I¹³¹) showed global hyperfixation within the enlarged gland (Fig. 3B). The perchlorate discharge test demonstrated a significant reduction (−84%) in radioisotope uptake after 1 hour, indicating a severe thyroid organification defect (Fig. 3C). Finally, knee X-ray confirmed the presence of distal femoral and proximal tibial ossification centers. Treatment was initiated with Levothyroxine at a dosage of 8.5 μg/kg/day. The patient's clinical follow-up has since been unremarkable, with successful stabilization of thyroid function and progressive reduction of goiter.

Figure 3.

(A) Thyroid gland ultrasound showing marked enlarged thyroid gland (white arrows). (B) Thyroid Scintigraphy (I¹³¹) showed global hyperfixation within the enlarged thyroid gland, before perchlorate administration. (C) Thyroid Scintigraphy (I¹³¹) after perchlorate administration. Perchlorate discharge test demonstrated a significant reduction in radioisotope uptake after 1 hour (−84%) (C), indicating a total thyroid organification defect. Created in BioRender. Stoupa, A. (2026).

Further discussion of patients #3 and #4

Dyshormonogenesis results from pathogenic variants in genes encoding components of the TH biosynthesis pathway, including SLC5A5 (NIS), TG, TPO, SLC26A4 (pendrin), DUOX2, DUOXA2, IYD (DEHAL1), and recently SLC26A7 (6, 36). Thyroid development and differentiation are normal, but one or more critical steps of hormone synthesis within thyrocytes are impaired, leading to hypothyroidism and, in some cases, goiter. Genotype-phenotype correlations have been established based on biochemical and imaging features (37, 38). Inheritance is typically autosomal recessive. Dyshormonogenesis is usually isolated. The two exceptions are Pendred syndrome (39-41), which is due to SLC26A4 mutations and combines goiter with sensorineural hearing loss and, in some cases, metabolic alkalosis; and SLC26A7 mutations (42, 43) with goitrous hypothyroidism and, in mice, tubular renal acidosis (44).

In patient #3, the neonatal goiter and sensorineural hearing loss warranted genetic investigations using the candidate gene panel approach with genes implicated in CH. Compound heterozygosity was demonstrated for SLC26A4 gene variants, namely, a missense variant (NM_000441.2, c.578C > T; p.Thr193Ile) (45) and another variant affecting the splice site (NM_000441.2, c.765 + 3A > T (46). Goiter in Pendred syndrome is very rarely neonatal and often develops only in late childhood or adolescence (47). There was no information on the iodine status of the mother or patient. In fact, iodine status may also have modified the phenotype of the patient and the presence of goiter in neonatal period.

In patient #4, molecular analysis performed by targeted gene panel sequencing identified two variants in the TPO gene: a missense variant (NM_001206744.2, c.617G > A, p.Arg206Gln)-already reported in the literature, without functional studies (48)- and a splice-site variant (NM_001206744.2, c.1598-1G > C, p.?) close to an acceptor splice site. In accordance with the American College of Medical Genetics and Genomics (ACMG) classification guidelines (49), both variants are classified as likely pathogenic.

Patients #5 and #6

Case reports

Two brothers born to nonconsanguineous parents were evaluated at our center for slow growth. Height was 178 cm (+0.3SD) in the father and 159.5 cm (−0.8SD) in the mother. There was no known family history of thyroid disease. No disorders were detected on the newborn screening panel.

The older brother (patient #5) was referred at 8.5 years of age. He was born at 36.5 weeks of gestation with birth weight and length appropriate for gestational age. Height and weight gains decelerated gradually from 4 years of age (height, mean to −1.5 SDS; weight, +1 SDS to −1 SDS; according to French growth charts). At referral, bone age was delayed by two years. A low fT4 level (9.52 pmol/L; normal, 12-22 pmol/L) in combination with a TSH level within the reference interval (2.26 mIU/L; normal, 0.27-4.2 mIU/L) suggested central hypothyroidism. Levothyroxine treatment normalized the TH levels. IGF-1 levels were also low (66 ug/L; normal, 95-240 ug/L). Growth hormone (GH) secretion as evaluated by a glucagon stimulation test was normal (peak, 31.8 mIU/L; normal, >20 mIU/L). The other pituitary axes were normal. MRI demonstrated marked anterior pituitary hypoplasia (2 mm); the stalk and posterior pituitary appeared normal. Growth velocity remained subnormal and IGF-1 levels were low, suggesting functional or partial somatotropic deficiency. Recombinant human growth hormone (rhGH) therapy started at 10 years of age resulted in statural catchup growth and IGF-1 normalization. After rhGH discontinuation at 13 to 14 years of age, growth remained normal, suggesting partial transient GH deficiency. Puberty was delayed, starting at 14.5 years of age. The pubertal growth spurt was normal, and final height was near the target height (177.5 cm, +0,2SD). Levothyroxine treatment was continued into adulthood, with dose adjustments and repeated assessments of thyrotropic function.

Given his brother's history, patient #6 was referred at 9 years of age for evaluation of growth deceleration. He was born at 38 weeks of gestation with birth weight and length appropriate for gestational age. His growth was linear at -1 SDS then decelerated to −1.5 SDS. Bone age at 3 years was delayed.

A low fT4 level (9.1 pmol/L; normal, 12-22 pmol/L) in combination with a TSH level within the reference interval (3.62 mIU/L; normal, 0.27-4.2 mIU/L) is consistent with central hypothyroidism. IGF-1 level was low-normal (126.9 ug/L; normal, 95-240 ug/L). Levothyroxine therapy was initiated. By MRI, the anterior pituitary size was normal (4.5 mm), as well as the stalk and olfactory bulbs. A GH stimulation test (glucagon-betaxolol) produced a normal GH peak (45.5 mIU/L). IGF-1 levels were closely monitored and remained normal. Under TH replacement alone, growth velocity stabilized without rhGH. At last follow-up at 13 years of age, the Tanner stage was 2, testosterone was 0.35 ng/mL, and LH was 0.42 IU/L, indicating incipient puberty.

Central congenital hypothyroidism

Permanent Central CH (CeCH) is due to hypothalamic-pituitary axis defects. In contrast to primary CH, central CH is not detected by TSH-based neonatal screening programs. Therefore, reported incidence figures depend on the neonatal screening approach. In several US states and in regions of Japan, where total T4 and TSH (US) and FT4 and TSH (Japan) are measured in the first week of life, reported incidences are around 1/106 304 (US) (50), and between 1/13 872 and 1/30 833, respectively (Japan) (51, 52). In the Netherlands, where a total T4, reflex-TSH, reflex-TBG strategy is used, the incidence is even higher: 1/16 404 to 1/13 000 (53-55).

In childhood or adolescence, growth deceleration, delayed pubertal development, or the emergence of additional pituitary hormone deficiencies may lead to the diagnosis.

CeCH is either isolated or, more often, associated with one or more other pituitary-hormone deficiencies (56). Morphological pituitary or hypothalamic abnormalities or other neurological defects may be present. The clinical picture varies widely depending on the underlying mechanism and genetic background. Typical signs of severe CH are often absent at birth. The diagnosis often relies on low or inappropriately normal TSH levels with low fT4 levels, particularly in children with growth failure, delayed puberty, or other pituitary hormone deficiencies.

Further discussion of patients #5 and #6

Given the presence of CeCH in two male siblings, targeted genetic testing was performed. A nonreported heterozygous pathogenic nonsense variant was identified in the IGSF1 gene (NM_001555, c.2170C > T; p.Gln724*) confirming X-linked IGSF1 deficiency syndrome. Despite sharing the same mutation, the brothers had different clinical manifestations. Both had CeCH, but only one had GH deficiency, pituitary hypoplasia, and delayed puberty (patient #5).

Table 1 lists the main genes involved in isolated CeCH. Isolated CeCH could be due to mutations in TSHB that cause defective TSHα/β heterodimerization with secretion of biologically inactive hormone, often with circulating TSHα elevation, a pathognomonic sign (57). However, pathogenic TSHB or TRHR variants are rare (58, 59).

Table 1.

Main genes implicated in isolated central hypothyroidism

Genes	Mode of inheritance	Severity of hypothyroidisma	Associated clinical features
TSHB (OMIM 188540)	AR	Severe	Reversible pituitary hyperplasia
TRHR (OMIM 188545)	AR	Moderate to mild If M/F carriers: recurrent high TSH levels	Low prolactine levels
IGSF1 (OMIM 300137)	X-linked	M: moderate F: low-normal FT4 to mild hypothyroidism	M: macroorchidism, delayed pubertal Testosterone rise F: delayed menarche M/F: GHD, low prolactine; increased BMI and fat mass
TBL1X (OMIM 300196)	X-linked	M: moderate to mild F: low-normal FT4 to mild hypothyroidism	hearing loss
IRS4 (OMIM 30904)	X-linked	M: mild F: low-normal FT4

^a Abbreviations: biallelic pathogenic variants, BMI, body mass index; F, females; GHD, growth hormone, deficiency; M, males.

X-linked IGSF1 deficiency has recently emerged as the major cause of isolated CeCH (55, 58, 60). Delayed or progressive thyrotrope dysfunction after the neonatal period is common. Transient partial GH deficiency, macroorchidism, variable prolactin deficiency, or delayed puberty may suggest the diagnosis (61). Similarly, pathogenic TBL1X variants have been identified in patients with isolated CeCH and sensorineural hearing loss related to impaired TH receptor-mediated transcriptional regulation (62). More recently, IRS4 mutations were found to cause isolated CeCH (63).

Combined CeCH typically results from mutations in genes encoding pituitary transcription factors and is associated with multiple pituitary hormone deficiencies and characteristic neuroimaging findings. PROP1 mutations are the most common cause of combined pituitary hormone deficiency and may result in progressive pituitary function impairment from childhood to adulthood, often with delayed-onset hypothyroidism. Mutations in developmental genes (eg, POU1F1, LHX3, LHX4, HESX1, SOX2, SOX3, LEPR, and OTX2) lead to variably combined deficiencies in GH, TSH, gonadotropin, and ACTH, with hypoplasia or midline brain defects (64, 65); however, most cases of pituitary stalk interruption syndrome are still unexplained. The CeCH is often clinically masked by the other manifestations of hypopituitarism.

In CeCH, genetic findings should be viewed primarily as providing diagnostic guidance rather than confirmation. A thorough clinical and biochemical evaluation, including tests for other pituitary hormone deficiencies, is crucial.

Advances and complexities in congenital-hypothyroidism genomics

Thyroid dysgenesis and the two-hit hypothesis

The pathogenesis of CH due to TD (CHTD) remains incompletely understood and probably involves an interplay between genetic susceptibility and environmental factors. Although most cases are sporadic, evidence for genetic causes (66, 67) exists: about 2% of patients with CHTD have an affected first-degree relative, the proportion is higher in consanguineous families and specific ethnic groups (3), and CHTD is strongly associated with extrathyroidal congenital anomalies and syndromic conditions (68, 69). However, the 92% discordance among monozygotic twins (70) and female predominance argue against classic Mendelian inheritance.

The two-hit hypothesis suggests that a predisposing germline mutation causes CHTD only when there is another somatic genetic or epigenetic defect in the thyroid gland or an adjacent structure (71). Regarding epigenetic defects, methylation profiles are not different between ectopic thyroid and leukocytes or orthotopic thyroid (72, 73). However, one study identified a difference in tissue-dependent FOXE1-promoter methylation between the thyroid and leukocytes (74). Also, a significant increase in hypomethylated stochastic epigenetic mutations in hypothyroid twin pairs compared to healthy twins has been reported (75). A genome-wide association study (GWAS) found a risk locus for thyroid aplasia or ectopia but not hypoplasia (76). This risk allele was associated with increased expression of FZD5 and CCNYL1, two mediators of Wnt signaling implicated in zebra fish thyroid development. This finding suggests that modulation of the genetic risk background may act as a second hit.

Oligogenicity

Oligogenic inheritance refers to the involvement of variants in two or more genes in a particular phenotype. A genetic variant in one gene may be insufficient to lead to a phenotype, whereas the combined effect of variants in two or more genes results in a specific phenotype. A role for oligogenicity in CH pathogenesis is suspected based on the higher incidence in families with at least one case of CH (67) and on the incomplete penetrance and variable expressivity in those families (77). Findings in a double heterozygous mouse model (Pax8/Nkx2.1) support this possibility (78). In this mouse model, thyroid gland defects occurred only with a specific genetic background (C57BL/6J mice). Other studies suggest an oligogenic mechanism in 21-26% of patients with CH (79, 77). This proportion may vary depending on patient selection, clinical phenotype, and variant classification (80, 81). Our group found a digenic mode of inheritance in 5.5% of 292 patients with CH and TD, each carrying a variant in DUOX2/DUOXA2 and in a thyroid developmental gene (82). In vitro functional studies and robust bioinformatic and statistical analyses support this model. CH may be caused by the addition of genetic variants to genes involved in thyroid development or hormonogenesis, with variable expressivity, penetrance, and functional defects explaining the sporadic occurrence and the phenotype variability within affected families.

Noncoding mutations

A noncoding variant responsible for dominantly inherited TSH resistance (RTSH) and multinodular goiter (MNG) has been identified (83, 84). GWAS or whole-genome sequencing detected a deletion or single-nucleotide variant in a noncoding short tandem repeat (STR) at 15q26.1 in TTTG₄. This STR is segregated with RTSH or MNG. It is among the 10% most constrained regions in the human genome and is located to a predicted cis-regulatory element with a thyroid-specific enhancer-like signature. Single-molecule chromatin fiber sequencing on thyroid tissue showed that the mutant STR activated a thyroid-specific enhancer cluster, leading to MIR7–2/MIR1179 upregulation (83). Small RNA sequencing confirmed microRNA overexpression of these two genes in thyroid tissue from affected patients. Overexpression of these microRNAs involved in modulating the PI3K/AKT/mTOR signaling pathway may impair thyrocyte proliferation and differentiation in an in vitro model (85).

Genetic approach and impact on patient management

A targeted genetic approach, guided by a detailed phenotypic description, has a direct impact on the management of patients with CH. According to 2020-2021 consensus guidelines, genetic counseling should not be routine but instead should be provided when indicated by the CH subtype, family history of CH, and presence of syndromic features (36). The American Academy of Pediatrics recommends genetic testing in CeCH and syndromic disease. In isolated primary CH, genetic testing is recommended only if the results would alter the clinical management (86).

In Fig. 4, we propose a genetic testing strategy for newly diagnosed patients with CH. When performed, genetic testing should follow a detailed evaluation of thyroid morphology and associated anomalies and should have a clear clinical objective, eg, to refine the diagnosis, inform the prognosis, guide the treatment, and/or assess the recurrence risk (Fig. 5). Modern genomic tools, including comparative genomic hybridization array and NGS (targeted gene panels, whole-exome/genome sequencing), are now preferred.

Figure 4.

Proposed genetic testing strategy for newly diagnosed patients with congenital hypothyroidism. Created in BioRender. NGUYEN QUOC, A. (2026).

Figure 5.

Genetic testing in congenital hypothyroidism: a precision-medicine approach based on diagnosis improvement, treatment targets, prognosis, and counseling. Created in BioRender. NGUYEN QUOC, A. (2026).

Accurate genotyping can identify the cause of the disease, clarify inheritance patterns, and predict thyroidal and extrathyroidal defect risks in relatives, particularly in syndromic CH, familial cases, and CeCH. Moreover, the genetic diagnosis may influence therapeutic decisions. For instance, in patients with pathogenic variants in genes implicated in TH synthesis (for example TPO or TG), individualized levothyroxine management prevents goiter formation, which can occur when the TSH level is not kept in the lower part of the normal range. Also, in patients with mild TSH resistance (monoallelic TSHR mutations), aggressive TSH normalization may be unnecessary (87) and patients with IYD variants can receive iodine supplementation instead of levothyroxine (88, 89).

The IYD gene encodes a dehalogenase enzyme that deiodinates monoiodotyrosine and diiodotyrosine residues generated during TG proteolysis. IYD thus enables intrathyroidal iodide recycling and contributes to efficient TH synthesis. IYD defect results in impaired iodide reutilization, depletion of TH precursors, chronic TSH stimulation, and goiter. When environmental iodine is scarce, a functional IYD gene is crucial for iodine conservation and utilization in TH production. In patients harboring biallelic pathogenic IYD variants, fT4 correlates positively with urinary iodine excretion, and the thyroid phenotype varies with the iodine intake (89). Iodine supplementation can be given instead of thyroxine supplementation (89, 88).

Genetic testing can help to distinguish permanent from transient CH in patients with DUOX2/DUOXA2 mutations, thus guiding treatment-withdrawal trials. However, both monoallelic and biallelic DUOX2 and DUOXA2 variants are associated with highly variable thyroid phenotypes, including delayed neonatal TSH elevation, TSH resistance phenotype, and either transient or permanent CH (90, 91). Therefore, the severity of thyroid dysfunction cannot be reliably predicted from the genotype alone, which should not be used to determine treatment duration. The phenotype variability is related not only to variant pathogenicity but also to iodine intake, developmental TH requirements (92), ethnicity (93), and, probably, genetic modifiers such as variants in genes encoding other H₂O₂-generating enzymes (eg, DUOX1) (94, 95).

Genetic testing identifies carriers of NKX2-1 mutations, who can experience life-threatening respiratory disease (Patient#1) (25). Pathogenic PAX8 variants are autosomal dominant or de novo and may be associated with urogenital tract anomalies (96, 97). FOXE1 pathogenic variants causing CH are autosomal recessive and responsible for Bamforth–Lazarus syndrome (thyroid agenesis, midline defects such as cleft palate, choanal atresia, and spiky hair). In this case, molecular diagnosis not only modifies the clinical management but also influences genetic counseling.

Finally, NGS has revealed roles for oligogenicity and noncoding variants and has also expanded genotype–phenotype relationships. These advances support a precision-medicine approach to CH, with genetics increasingly informing counseling, treatment targets, follow-up strategies, and anticipatory guidance for patients and families (Fig. 5).

Pitfalls and limitations

NGS is now more widely available and relatively affordable. Nonetheless, several pitfalls should be considered. First, many variants are classified as variants of uncertain significance, particularly those in genes with incomplete penetrance. Variants should be interpreted according to the thyroidal and extrathyroidal phenotype, familial segregation studies, and comprehensive biological screening of index cases and relatives. Moreover, variant interpretation is intrinsically linked to the current state of knowledge. Therefore, reanalysis of exome/genome datasets initially considered negative is warranted, particularly as novel gene-disease associations and annotation frameworks emerge over time.

Second, incomplete penetrance or variable expressivity is not rare with genes causing CH (eg, PAX8 and genes encoding other developmental transcription factors). Thus, phenotypes may vary within the same family (98, 99). Consequently, the absence of CH in variant-carrying relatives does not exclude pathogenicity, and the presence of a variant does not always result in disease.

Third, widely used genetic diagnostic tools (comparative genomic hybridization array, targeted gene panels, and exome sequencing) may fail to detect pathogenic mechanisms outside coding regions, such as deep intronic variants, structural variants, or copy-number variations (100). Even the most recent NGS technology has disadvantages, including the need for nonstandard DNA extraction and the absence of sufficiently large and diverse control cohorts. Indeed, many ethnicities are not represented or are underrepresented, making the interpretation of DNA variants in these populations more difficult (101). Usual genomic technologies cannot assess molecular mechanisms identified by transcriptomic analysis (eg, differential expression, gene fusion, alternative splicing, and RNA editing) or by epigenomic methods (eg, DNA methylation, histone modifications, and transcription factor binding) (72, 73, 83). Similarly, the effects of oligogenic inheritance (77, 82), and endocrine disruptors (102) on genes are difficult to elucidate using standard monogenic interpretation frameworks.

Finally, technical and ethical issues include the nonstandardized gene panel content (exon coverage, sequencing of untranslated regions, or even genes), differences in bioinformatic pipelines, and the psychological impact of uncertain or incidental findings in neonates.

Conclusions

Despite major advances in CH genetics, the immediate impact of a molecular diagnosis on the initial management remains limited. Neonatal treatment decisions rely primarily on biochemical severity and clinical context. Genetic findings rarely modify decisions about early levothyroxine initiation or dosage. Early neurodevelopmental outcomes depend chiefly on an early diagnosis and appropriate levothyroxine supplementation rather than on determining the genotype.

Genetics plays an important role beyond the neonatal period. The molecular diagnosis can help optimize long-term replacement therapy. Individualization may be improved by combining algorithm-based approaches and a genetic diagnosis (103).

The genetic diagnosis helps to determine the prognosis, notably in familial, syndromic, and central forms of CH. Genetic information guides long-term follow-up strategies, including the intensity of biological monitoring, appropriateness and timing of treatment withdrawal trials, and anticipation of associated thyroidal or extrathyroidal manifestations.

Importantly, the benefits of targeted genetic testing are greatest when effective neonatal screening is performed. At present, only about 30% of neonates worldwide are screened for CH (1). Every effort should be made to increase this proportion and strengthen public health strategies in this direction.

Learning objectives

Upon completion of this educational activity, participants should:

• be familiar with the clinical management of patients with congenital hypothyroidism (CH);

• be able to develop a step-by-step diagnostic-workup plan including consideration of genetic tests;

• know how genetic information guides clinical decision-making; and

• be familiar with current controversies about using genetic testing to screen neonates for CH.

Disclosures

None of the authors has any conflicts of interest to disclose relative to the content of this article.

Data availability

Data sharing is not applicable to this article, as no datasets were generated or analyzed for the current work.

Authors' declaration

The authors affirm that the work submitted for publication is original and has not been published in any language or format and has not been submitted elsewhere for print or electronic publication consideration. Each person listed as an author participated in the work in a substantive manner, in accordance with ICMJE authorship guidelines, and is prepared to take public responsibility for it. No AI or machine-learning tools were used to prepare the manuscript. All authors consent to the investigation of any improprieties that may be alleged regarding the work. Each author further releases and holds harmless the Endocrine Society from any claim or liability that may arise therefrom.