Newborn screening for congenital hypothyroidism (CH) enables prevention of neurodevelopmental delays and achievement of an intellectual quotient in the normal range in affected patients.1,2 In most autonomous communities in Spain, one problem remains: since screening is limited to measurement of thyrotropin (TSH), it fails to detect cases of central hypothyroidism—of hypothalamic-pituitary origin—in which TSH levels are normal or only mildly elevated.2 Historically, this type of CH has been considered rare and mild, so its inclusion in screening programs did not seem essential. We now know that it is not as rare as previously believed, as it is present in 1 per 13 000 to 1 per 16 000 live births, nor as mild, since neurological impairment of varying severity occur in half of the cases. Scientific societies, both in Spain and abroad, now recommend screening with simultaneous measurement of thyroxine (T4) and TSH in the dry blood spot sample.1,2
We present the cases of two brothers aged 4 and 2 years with an unremarkable medical history, except for more severe neurodevelopmental delay in the older sibling, who exhibited motor impairment and learning difficulties. The patient was referred to us after confirmation of low free thyroxine (T4) levels associated with inappropriately normal TSH. The rest of the pituitary hormone levels were normal, except for low prolactin. Treatment with levothyroxine was initiated, and after obtaining a normal neuroimaging result, laboratory testing was ordered for his brother, who had a language delay, which identified the same hormone abnormalities.
The genetic study (whole exome sequencing with Twist Human Core Exome using the Illumina NextSeq 1000 sequencing system) identified a heterozygous variant (c.3482dup; p.Lys1162*) in the immunoglobulin superfamily member 1 (IGSF1) gene in both siblings and the mother, a variant that, although not registered in the Clin Var database, was interpreted as likely pathogenic because it was a stop-gain change. The mother was healthy, had not had problems breastfeeding and had free T4 and prolactin levels in the lower limit of normal. Table 1 presents the clinical data for the family.
Clinical and laboratory characteristics of the patients and the mother.
| Clinical characteristics at diagnosis | Sibling 1 | Sibling 2 | Mother |
|---|---|---|---|
| Age in years | 4.0 | 1.84 | 33 |
| Height cm/z | 102.0 /−0.5 | 86.0 /−0.6 | 160/−0.69 |
| Weight kg/z | 22.0/+1.77 | 15.4/+1.64 | 55.3/−0.27 |
| BMI kg/m2/z | 25.1/+2.94 | 20.8/+2.96 | 21.6/+0.08 |
| Presentation | Intellectual disability | Language delay | None |
| General examination | Normal | Normal | Normal |
| Neurologic examination | Normal | Normal | Normal |
| Signs of hypothyroidism | None | None | None |
| Hearing | Normal | Normal | Normal |
| Clinical characteristics at birth | |||
| Weight g (z) | 3670 (+0,93) | 3900 (+1,26) | |
| Length cm (z) | 53,0 (+1,51) | 53,5 (+1,88) | |
| Test results (normal range) | |||
| Free T4 ng/dL (0.8−1.8) | 0.55 | 0.65 | 0.85 |
| Free T3 pg/mL (1.58−3.91) | 2.43 | 3.01 | 2.91 |
| Thyrotropin μU/mL (0.7−6.6) | 6.60 | 5.03 | 1.73 |
| Prolactin ng/mL (4.04−15.20) | 2.7 | 1.2 | 5.0 |
| Cortisol μg/dL (6.5−22.7) | 18 | 19 | 19 |
| IGSF1 ng/mL (12−120) | 75 | 53 | 130 |
| Total cholesterol mg/dL (<200) | 241 | 208 | 189 |
| LDL-cholesterol mg/dL (<130) | 182 | 139 | 87 |
| HDL-cholesterol mg/dL (>40) | 56 | 57 | 85 |
| Triglycerides mg/dL (<90) | 69 | 62 | 86 |
| MRI of the sella turcica | Normal | Not performed | Not performed |
BMI, body mass index; HDL, high-density lipoprotein; IGSF1, immunoglobulin superfamily member 1 gene; LDL, low-density lipoprotein; T3, triiodothyronine; T4, thyroxine; z, z-score.
The most frequent cause of isolated central CH is IGSF1 variants, responsible for 40% of cases. Two deletions and two mutations have been reported to date, which in male carriers cause X-linked CH in every case, hypoprolactinemia in two thirds of cases and other endocrine abnormalities with lesser frequency, such as growth hormone deficiency, obesity, delayed adrenarche, slow progression of puberty (with onset at normal age and normal gonadal function in adulthood) and macroorchidism. Women who carry IGSF1 variants tend to be asymptomatic and have free T4 and prolactin levels around the lower limit of normal. In some cases, female carriers develop central CH (18%–33%), delayed menarche, or overweight.3,4
As far as we can tell, there are no previous reports of the variant identified in this family. Given that the clinical phenotype and laboratory features of its members were fully consistent with those previously described in association with IGSF1 defects, it is reasonable to conclude that our finding expands the variant spectrum of this gene
Patients with central CH born in areas where newborn screening does not include measurement of T4 are diagnosed later in childhood due to neurodevelopmental delays and short stature associated with hearing loss and hypercholesterolemia. There is a spectrum of disease severity, and it is directly correlated to age at diagnosis and inversely correlated to thyroid hormone levels, as is the case in primary CH when screening is not performed. In cases of central hypothyroidism, since thyroid function is partially preserved, neurodevelopmental delay is usually not severe and may present in isolation, without any of the typical manifestations of hypothyroidism, as these cases illustrate. The areas of neurologic functioning most likely to be affected are language, attention, and mental activity; therefore, in cases of mild and moderate severity, patients may present only with language delay, attention difficulties, and mental fatigue. However, there have also been reports of florid cases of central CH caused by IGSF1 variants with severe intellectual disability and short stature associated with jaundice, hearing loss and the other typical features of CH.4
As is already done in the Netherlands, Israel and some regions in other Western countries,5 including the Basque Country and Cantabria in Spain,6 it would be very beneficial to include measurement of T4 levels in the newborn screening program to allow early diagnosis and treatment of central hyperthyroidism in order to prevent the development of potentially severe neurologic impairment. Until this measure is implemented, clinicians should assess thyroid function (including free T4 levels) as soon as possible in all children presenting with any kind of neurodevelopmental abnormality.



