Triiodithyronine (T₃) is the biologically active form of TH but it is important to remember that as T₃ cannot pass the brain barriers in significant quantities, most T₃ in the brain is converted from local deiodination of T₄, meaning that in CH patients, it is not necessary to replace T₃ to normalise neurologic development (1). Research involving 47 infants treated with different doses of l-T₄, found that serum T₃ normalized and stayed normal regardless of dosage, indicating that l-T₄ alone is an adequate treatment for CH (30).

It is important to start treatment promptly in all infants who test positive in CH screening, after samples for confirmatory tests are taken, the results of which will determine the optimal l-T₄ starting dose (31). When the results of venous blood tests are not available on the same day, treatment should be started without waiting for the results, particularly if the patient’s condition is a cause for concern (30, 32, 33),

The treatment goals for CH are outlined by the American Academy of Pediatrics (AAP) (30) and by the European Society for Paediatric Endocrinology (ESPE) (20) and can be summarized in the following main points:

Some data suggests that infants with serum T₄ levels under 10 μg/dL during the first 12 months of life, associated with serum TSH levels greater than 15 mU/L, have a lower Intelligence Quotient (IQ) than those with serum T₄ levels above 10 μg/dL (28). Furthermore, other data show that treatment with higher doses of l-T₄ may be connected with higher IQ at 7 and 8 years of life, particularly in verbal working memory and sentence comprehension (27). Frequent and significant alterations in serum T₄ and TSH levels during the first 12 months of life are associated with differences in mental development index scores and verbal IQ (27, 33).

Prompt diagnosis and initiation of therapy for CH is essential for ensuring positive outcomes, especially in high-risk infants (34). However, some patients with CH may not achieve normal TSH concentrations despite l-T₄ treatment even at higher doses. Poor adherence to treatment is the most frequent reason for poor TSH control. Therefore, in patients who do not respond to treatment it is important to investigate whether l-T₄ is being correctly administered and if necessary, review the method of administration. The possibility of impaired l-T₄ absorption related to gastrointestinal diseases or the interference of drugs or other substances (see below) should also be considered (35).

Good therapeutic management practices appear to be of particular importance to ensure optimal results in terms of neuromotor and mental development, due to the THs-dependent processes of brain development in the first years of life. Professionals should provide clear explanations and instructions to caregivers to improve adherence to treatment (9, 34, 35). It is critical that families comprehend that treatment is essential for the best patient outcomes.

Infants with severe forms of CH are at a higher risk for ID; for this reason, the dosage and timing of l-T₄ treatment are fundamental for optimising neurocognitive outcomes in the future. Delays in normalizing serum T₄ of more than one week can lead to lower IQ (23). Some data suggest that scores in behavioural and cognitive tests were lower for patients whose T₄ did not normalize before two weeks than for patients who achieved normal levels in under two weeks (24). Therefore, as stated above, the treatment goal is to bring serum T₄ to above below 129 mmol/L (> 10 μg/dL) as quickly as possible. The data available in the literature show that there is an inverse relationship between the initial l-T₄ dose, and the time taken to reach target serum T₄ levels (24) meaning that the prompt administration of an appropriate starting dose of l-T₄ is particularly important.

The starting l-T₄ dose recommended both by AAP and the ESPE is from 10 to 15 µg/kg/day (19–21, 30). In infants born at term this translates to an average starting dose of 37.5 - 50 µg per day (20, 21). In the above cited research, infants receiving 50 µg per day (corresponding to 12 - 17 µg/kg/day), versus 37.5 µg (corresponding to 10 - 15 µg/kg/day) performed better in tests measuring behaviour, reading, spelling and maths. The patients on the higher dose achieved IQ scores 11 points higher than those on the lower dose (24). Therefore, in infants whose serum T₄ levels are lower than 5 μg/dL, higher l-T₄ doses are recommended.

The goal is to normalize T₄ and TSH within 2 and 4 weeks, respectively (19, 24). T₄ and FT₄ concentrations should be in the upper half of the reference range for age, with normalization of TSH (19).

These positive outcomes associated with administering higher doses of l-T₄ have been highlighted by studies in Europe, the United States and Canada (24, 36–38). In one study, 83 infants were placed into three different groups, each receiving different initial doses of l-T₄ from diagnosis at birth (the first group was treated with 6.0 - 8.0 µg/kg/day, the second with 8.1 - 10.0 µg/kg/day, and the third with 10.1 - 15.0 µg/kg/day) and followed up at four years of age for growth and intellectual development. The infants with severe CH who started on the highest dose achieved the highest intellectual scores (39). Another study involving 61 infants comparing early treatment with delayed treatment and low doses with high doses, found that the outcomes for patients with severe forms of CH were more favourable in those treated early (under 2 weeks) and with higher doses (>9.5 µg/kg/day). Such patients presented normal psychomotor development at 10-30 months of age (23), showing that the time it takes for TSH to normalize is inversely correlated to neurological outcome (19).

Another study which analysed data on the intellectual development of 45 CH children found that differences and variations in serum levels of T₄ and TSH during the first year of treatment predicted scores on mental development index (MDI) and verbal IQ tests performed at 2 years and 6 years of age (33). It is clearly imperative to monitor patients so that l-T₄ doses can be promptly adjusted and optimal T₄ and TSH levels maintained.

CH treatment must be carefully monitored with appropriate adjustment of the l-T₄ dose, until serum TSH and FT₄ concentrations have normalized. Clinical evaluations must be made every few months during the first three years of life together with frequent measurements of serum T₄ or FT₄ and TSH, in order to avoid prolonged periods of under- or overtreatment (19, 24). Management should focus on maintaining a status of euthyroidism, especially in the first 3 years of life when brain development is most dependent on THs. However, in CPHD patients with associated adrenal insufficiency or in patients in whom this problem cannot be excluded, adrenal function should be investigated, and glucocorticoid therapy should precede l-T₄, replacement treatment to avoid inducing an adrenal crisis (19).

The AAP recommends monitoring thyroid function 2 and 4 weeks after initiating l-T₄ treatment and every 1 to 2 months during the first 6 months of age. After this, patients should be monitored every 3-4 months between 6 months and three years of age, and then every 6-12 months until the child reaches adult height. Moreover, TH should be evaluated four weeks after any change in dose and follow-up evaluations need to be more frequent if results are abnormal or non-adherence to therapy is suspected (19).

Although in most infants, serum T₄ will normalize within seven to fourteen days and serum TSH will normalize within one month of treatment, some patients continue to present altered TSH levels (10-20 mU/L) or serum T₄ levels beyond these timeframes (40). This is often caused by under treatment, although it is important to remember that in 10% of CH infants undergoing treatment the FT₄ feedback control on TSH secretion does not mature normally (41) due to the resetting of the pituitary-thyroid feedback mechanism in utero (42). In one paper reporting data for 42 CH paediatric patients, a prevalence of pituitary TH resistance of 43% in patients below one year of age was found. However, in childhood and adolescence this problem was found in only 10% of patients (42), suggesting that TH resistance is more common in younger patients and may decrease with age.

The availability of adequate TH is essential for normal brain development in the first two to three years of life. Low TH levels during this period can lead to irreversible damage, whereas after 3 years of age the effects of hypothyroidism can usually be reversed when the condition is treated. A study by the New England Congenital Hypothyroidism Collaborative highlighted the importance of regular monitoring and dose adjustments to control serum FT₄ and T₄ levels. Researchers found that eighteen infants included in the study with reduced serum T₄ levels, who were taking insufficient doses of l-T₄ (< 5 µg/kg/day) and who had an anamnesis showing poor compliance in the first three years of life, had a mean IQ of 87 (43). In comparison a larger group of correctly treated patients, with a history of normal serum T₄ levels, achieved a mean IQ score of 105.

Further data suggest that noncompliance after the first three years of life may also negatively affect cognitive achievement. For example, researchers made unannounced home visits on 14-year-old adolescents with CH and found that 44% were not adequately treated, having serum TSH > 15 mU/L and T₄ levels < 6.6 µg/dL in the majority (44). Psychometric tests revealed a mean IQ of 106. The importance of complying with TH treatment was explained to patients and their families (the dose was not altered), and 12 to 24 months later, serum TH levels had improved, and the repetition of psychometric tests showed an improvement in mean IQ of 6 points. Although the study did not include a control group, the authors concluded that non-compliance during adolescence is frequent, and that improved compliance can lead to improved cognitive function.

ID caused by CH is preventable but lack of awareness amongst caregivers about the benefits of treatment may be a barrier to clinical improvement in some patients (7). Newborn screening represents a unique opportunity for avoiding damage connected to CH (7).

The effects of overtreatment with l-T₄ are well known (45–47). A prospective study by Rovet et al. (27) demonstrated for the first time that, despite an improvement in cognitive profiles, an initial high dose of l-T₄ can be associated with behavioural problems in childhood. Bongers-Schokking et al. (48) evaluated DQ scores at 11 years, showing that overtreatment can lead to a cognitive impairment of −17.8 points if it is prolonged and −13.4 points when the overtreatment is limited to the first 2 years of life. Another recent study by Bongers-Schokking et al. (29) concluded that episodes of overtreatment during the first 3 months can lead to permanent ADHD, while undertreatment may be related to autism.

l-T₄ is available in several forms (tablets, soft gel capsules, liquid and, in some countries, powder for preparing an intravenous solution). Unfortunately, tablets are the only form available in many countries.

It is clear that interference from food can negatively impact the efficacy, and in some cases the safety, of treatment with l-T₄. This is also the case for liquid and soft gel capsule formulations (101). Most data, especially sporadic case reports, focus on adults and there is a need to augment our knowledge about the effects of food on l-T₄ in children. In the presence of problems of l-T₄ absorption, switching from tablets to liquid l-T₄ or soft gel capsules can in part solve the problem, and observing appropriate intervals between taking l-T₄ and eating can also reduce the possibility of interaction (105).

Many conditions or drugs may influence the gastric environment and affect the absorption of l-T₄. Tablets of l-T₄ need to be dissolved in an acid environment. Subjects with achlorhydria, reduced gastric acidity or who are taking PPIs do not have the ideal gastric environment to dissolve l-T₄. PPIs are increasingly prescribed to children (122) and are effective treatments for many gastric diseases, duodenal ulcers, non-steroidal anti-inflammatory-induced ulcer-related prophylaxis, Helicobacter pylori in combination with other medications, gastro-esophageal reflux disease and eosinophilic esophagitis. They are also prescribed for functional dyspepsia, chronic cough, and infantile reflux (122).

In adult goitrous patients taking l-T_4, omeprazole assumption led to significant increases in serum TSH; this effect was reversed by increasing the dose of l-T₄ or by discontinuing the use of omeprazole (123). Similar results have been reported for lansoprazole, but not pantoprazole or esomeprazole taken for one week (57). Antiacids also appear to interfere with thyroxine absorption by influencing gastric pH (57).

The recently developed liquid l-T_4, dissolved in an alcoholic solution, does not need an acid environment as it can be directly absorbed through the intestinal mucosa (124).

Besides drugs, pre-existing malabsorption diseases or disorders that impair gastric acidity also affect the bioavailability of l-T₄ (125). There are some reports of elevated serum TSH levels in patients with coeliac disease and inflammatory bowel diseases taking doses of thyroxine that were previously sufficient to normalise serum levels (125).

Some studies report similar findings for patients with Helicobacter pylori infection and atrophic gastritis – both of also which impair gastric acidity (125).

Several reports regarding cases of l-T₄ malabsorption in patients with various manifestations of CD have been reported (125, 126). For example, in a study conducted on seventy-nine children with permanent CH, 6 patients (4 girls, 2 boys) were positive for CD antibodies, showing a higher-than-normal prevalence of this disease in children with permanent CH. The available data show that a gluten-free diet results in the reduction of l-T₄ dose requirements (127). Because the symptoms of some forms of CD are subtle, several authors suggest screening with CD markers in patients with hypothyroidism who need higher than normally expected doses of l-T₄.

Franzese et al. report an infant with CH who developed, during l-T₄ replacement therapy, a cow’s milk protein intolerance and subsequently CD. Both milk protein intolerance and CD affect the intestinal absorption of l-T₄, making the management of CH difficult. After starting a diet with a hydrolyzed milk protein and maintaining the dose of l-T₄ at 12 µg/kg dose, serum T₄ improved and TSH decreased (128). A case of l-T₄ malabsorption with lactose intolerance was reported in a 55-year-old woman with primary hypothyroidism whose TSH levels were persistently high (128).

After being diagnosed with oligo-symptomatic lactose intolerance, the patient was given a lactose-free formulation of l-T₄ (150 mg daily) and began following a lactose-restricted diet; after 3 months the problem had resolved (129).

The presence of a pylori infection and atrophic gastritis of the body of the stomach, causing bacterial production of urease that neutralises gastric pH, may determine a decreased TSH suppression (123).

An infant with confirmed CH taking l-T₄ experienced a possible drug interaction with simeticone (130) which is used to treat infant colic, a condition frequently seen by paediatricians. In this case, despite adequate l-T₄ dosage, TSH was high, suggesting undertreatment. This case highlights the importance of regularly reviewing a patient’s clinical history and considering the possibility of drug interactions in unusual circumstances, especially where patients need high doses despite good compliance and proper administration (131). Clinicians should alert the parents of CH children starting l-T₄ about the use of medicines containing simeticone (130).

Small intestinal bacterial overgrowth (SIBO) is a heterogeneous condition with nonspecific symptoms characterized by the presence or increase of atypical bacteria in the small intestine that unbalances intestinal microbiota. SIBO is characterized by symptoms such as flatulence, distension and abdominal pain. These symptoms might be indistinguishable from those presented by patients with functional gastrointestinal disorders. In most cases, SIBO is associated with motility or inflammatory disorders (132). The abnormally high levels of bacteria in the small intestine can cause malabsorption and suboptimal responses to narrow therapeutic index medication which are absorbed in the small intestine, such as l-T₄ (133). In one study on patients with SIBO, l-T₄ tablets and a compounded oral suspension were not absorbed efficiently, resulting in below optimal control of serum TSH. When patients switched to l-T₄ sodium oral solution, TSH levels fell and symptoms resolved (133).

Malabsorption of l-T₄ has also been described in patients infected with Giardia lamblia. Giardiasis is a common intestinal infection, but it has only rarely been reported as a cause of impaired l-T₄ absorption (124). Giardiasis is probably under-diagnosed; in developing countries with poor sanitation its prevalence is about 20% while in industrialised countries prevalence ranges from 3% to 7% (124). Intestinal giardiasis can cause maldigestion, malabsorption and diarrhoea, leading to anaemia, weight loss and growth retardation (124). In cases of giardiasis infection in patients taking l-T₄, switching from tablets to oral solutions revert decreased l-T₄ absorption (124).

Short bowel syndrome (SBS), a major cause of intestinal failure in children, is diagnosed when the length of the small intestine is 25% shorter than expected for the child’s age (131). In paediatrics, the most frequent causes of SBS are resections secondary to necrotizing enterocolitis, gastroschisis, intestinal atresia and intestinal volvulus (134). SBS can cause a number of metabolic changes in the body, which can affect growth, intestinal adaptation and lead to metabolic bone disease as well as an impaired functioning of the hypothalamic-pituitary-thyroid (HPT) axis (135). Passos et al. reported six consecutive cases of children with SBS with associated hypothyroidism during the period of intestinal rehabilitation (135).

A fibre-rich diet or dietary fibre supplements may also significantly reduce the bioavailability of l-T₄ due to a non-specific link to fibre in the intestinal lumen causing malabsorption of the hormone (125). Products containing insoluble dietary fibre increase bowel motility, potentially altering the intestinal absorption of l-T₄ (78). Although not all authors agree (135), in cases of dietary modifications which increase fibre intake, it may be necessary to monitor TSH levels and increase the dose of l-T₄ (136).

The possibility of interactions between levothyroxine and essential and trace elements has been investigated. Di- and trivalent elements, especially calcium and iron, appear to decrease l-T₄ bioavailability (137). The mechanisms involved are not entirely clear but may connected to unspecific adsorption and the formation of insoluble complexes in the intestine (138). Parents should thus be informed about the possible adverse effects of calcium and iron supplementation.

It has been shown that increased gastric pH can influence the absorption of l-T₄. Jubiz et al. studied the effects of vitamin C, capable of lowering gastric pH, on l-T₄ absorption (170) They conducted a study in 31 hypothyroid subjects with gastritis under l-T₄ therapy, taking a median dose of 100 µg in tablet form, with 120 mL of water with or without 500 mg vitamin C in solution. While on vitamin C, TSH levels fell in all patients, and FT₃ and FT₄ levels increased significantly. Antúnez et al. (171) observed similar results in 28 patients with elevated TSH levels, despite being on a l-T₄ dose higher than 1.70 µg/kg. Patients were asked to take l-T₄ tablets with 1 g of vitamin C (effervescent tablets, dissolved in 200 mL of water) for 6–8 weeks. Significant decreases in serum TSH levels (from 9.01 ± 5.51 mU/L to 2.27 ± 1.61 mU/L) were achieved.

Biotin is a vitamin commonly used in multivitamin preparations and used to treat progressive multiple sclerosis, several inherited metabolic diseases and acquired dermatologic diseases. It can interfere with many endocrine laboratory assays (including TSH and FT₄) (172), which can lead to erroneous or delayed diagnosis of CH. There are cases in the literature of children taking high doses of biotin as therapy for metabolic disease being wrongly diagnosed with hyperthyroidism (172).

The correct assessment of thyroid function may be influenced by biotin therapy (173). High levels of biotin–streptavidin can compete with biotinylated components leading to inaccurate and misleading results characterised by increased FT₃, FT₄, levels and a reduction in TSH levels, which may lead to misplaced suspicion of Graves’ disease (173). This could lead to erroneous therapeutic strategies with serious consequences in the first three years of life given the thyroid dependency of the central nervous system in this period.

Wijeratne et al. conclude that interference peaks at approximately 2 h after biotin ingestion and can persist for up to 24 h (174). In a newborn with trisomy 21, CH and partial biotinidase deficiency, because of interference with the TSH assay from concurrent biotin administration routine screening detected “normal” TSH levels leading to a delayed diagnosis of CH (172).

In paediatrics, the most basic and important method of nutritional intervention is enteral nutrition (EN) (175). EN is introduced for patients who cannot meet their energy and nutritional needs through normal feeding. EN is often required for children with growth retardation, inadequate weight gain, or weight deficit, and is employed in the treatment of diseases such as Crohn’s, and food allergies or intolerance (175). In neonates, EN is required in situations such as premature or necrotizing enterocolitis (175).

Reis et al. (176) carried a multicentre study in Brazil to evaluate possible drug-EN interactions. The authors found l-T₄-EN interaction to be one of the most frequent and clinically significant interactions. In an earlier study, Dickerson et al. (177) evaluated 13 hypothyroid patients in hospital: all participants received EN; their l-T₄ doses were kept the same as they were before the patients were hospitalised for 20 ± 5 days. Eight of the thirteen patients developed subclinical or overt hypothyroidism, indicating that it is vital to monitor hypothyroid patients who receive continuous EN and who are being treated with l-T₄. Manessi et al. (178) discovered that l-T₄, may be adsorbed by enteral feeding tubes and hypothesised that this mechanism was related to a reduction in treatment efficacy. However, Wohlt et al. (179) concluded that the amount of l-T₄ absorbed by feeding tubes is likely to be clinically irrelevant, and that impaired l-T₄ absorption may be caused by the concomitant ingestion of food. Pirola et al. (180) studied 20 euthyroid patients, a day after surgical invention, to compare the efficacy of different formulations of l-T₄ administered while patients were attached to an enteral feeding tube. EN was stopped for 30 min before and after powdered l-T₄ tablets were administered, whereas liquid l-T₄ was administered via the feeding tube without interrupting EN. There were no significant differences in the results obtained by the two methods, leading the authors to conclude that EN does not impact the absorption of liquid l-T₄.

Parenteral Nutrition (PN) is used when normal and enteral feeding are impossible (181). Some reports show that long-term PN is a risk factor for low iodine and/or selenium levels, causing hypothyroidism, which in some cases is severe, stressing the importance of dosing of these elements and evaluating THs (182, 183).

Glucocorticoids are often used in paediatrics for their immunosuppressant and anti-inflammatory properties, individually or in combination with other drugs both short and long term (186, 187). It is well known that GCs affect serum TSH levels in animals and humans, reducing TSH secretion by directing effecting TRH a hypothalamic level (188). Physiologically, it appears that levels of hydrocortisone have a key role in the diurnal variation of serum TSH, with lower levels in the morning and higher levels at night (189, 190). Some data have shown that high dose GCs may suppress serum TSH in normal subjects and also in hypothyroid patients (191). Nevertheless, long-term high doses of GCs or the cortisol excess typical of Cushing’s syndrome do not seem to result in clinical central hypothyroidism necessitating l-T₄ treatment (192). Dexamethasone doses as low as 0.5 mg or prednisone 30 mg can reduce TSH levels significantly (192).

This suppressive effect of GCs on TSH secretion appears to be controlled by the hypothalamus through the inhibition of TRH (193). In humans, GCs receptors are present in the TRH neurons of the paraventricular nucleus and high dose GCs can reduce TRH mRNA levels in the hypothalamus, acting on the TRH gene, probably the primary mechanism for lowering TSH secretion from the hypophysis (194).

The effect of GCs on TSH should be taken into consideration when checking thyroid function in CH patients, in order not to mistake an overtreated condition with a physiological response to GC treatment.

Dopamine is widely used in critical illnesses to increase blood pressure, cardiac output, urine output, and peripheral perfusion in neonates, infants, and older children with shock and cardiac failure (195).

Dopamine and dopamine agonists can suppress serum TSH and affect the activation of dopamine D2 receptors, reducing TSH pulse amplitude without significantly altering TSH pulse frequency (196, 197). Interestingly, in rats, dopamine stimulates hypothalamic TRH release, but because its overall effect is to lower serum TSH, the inhibitory effect on the pituitary exceeds the first effect (198).

Data on neonates with nonthyroidal illness (NTI) syndrome treated with dopamine infusions indicate that dopamine and NTI have an additive effect on suppressing the HPT axis, placing these individuals at risk of iatrogenic central hypothyroidism (199, 200). It is not clear whether treatment with l-T₄ should be given to patients with NTI who are receiving dopamine infusions but, given the thyroid dependence of the central nervous system, the risks of not treating children with CH under 3 years of age should be carefully considered.

Several drugs used to treat epilepsy such as carbemazepine (CBZ), oxcarbemazepine and valproic acid (VPA) could increase the metabolism of TH through the hepatic P450 system and could also impede pituitary responsiveness to hormonal feedback and cause central hypothyroidism (201, 202). However, several investigations suggest that the hypothalamic-pituitary axis is unaffected by these drugs and a specific mechanism has not been discovered (203). Effects appear to be different in relation to the medications considered. For example, in a study evaluating the use of VPA, CBZ and levetiracetam (LEV), thyroid dysfunction was frequent in children taking VPA or CBZ as a monotherapy, but absent in those taking LEV. The authors suggest that, in children with a predisposition for thyroid disease, LEV should be considered over VPA and CBZ, if appropriate for seizure and epilepsy type (204).