Increased glucose utilisation

The most common cause of excessive utilisation of glucose is neonatal hyperinsulinism. Hyperinsulinism should be confirmed by the use of a highly specific insulin assay for plasma insulin concentrations and its interpretation with reference to normal neonatal insulin–glucose relationships ( ) ( Fig. 34.3 ). Clinical features are that glucose requirements to maintain normoglycaemia are high, in excess of 8 mg/kg/min, as compared with 4–6 mg/kg/min usually required by neonates, and the infant may be macrosomic if hyperinsulinism was of fetal origin ( Fig. 34.4 ). Investigation of suspected hyperinsulinism will demonstrate low fatty acid and ketone body concentrations during hypoglycaemia, but this feature is not specific to hyperinsulinism as some infants who are not hyperinsulinaemic, such as some who are preterm or have IUGR, also fail to mount lipolytic and ketogenic responses.

Insulin–glucose relationship in preterm neonates; insulin concentrations measured using a highly specific immunoradiometric assay.

(Reproduced from .)

Newborn infant with hyperinsulinism showing increased adiposity and resemblance to an infant of a diabetic mother.

Insufficient supply of glucose

Hypoglycaemia is most often the result of reduced delivery of glucose into the blood. In the enterally fed infant all the circulating glucose is provided either by the absorption and conversion of sugars or by glycogenolysis and gluconeogenesis, and in some babies there may be a contribution from intravenous glucose infusion. Thus, if the infant fails to switch on glycogenolysis or gluconeogenesis in response to falling blood glucose levels, or clinicians prescribe insufficient intravenous glucose, hypoglycaemia may occur. This will be most significant when production of alternative fuels to glucose is also impaired. Three possible mechanisms may cause the failure of glucose production: (1) reduced availability of gluconeogenic precursors; (2) reduced activity of enzymes of glycogenolysis and gluconeogenesis; or (3) impaired counterregulatory hormone response.

Normal babies

As described above, healthy full-term AGA neonates often have low blood glucose concentrations in the first postnatal days, but are protected by the presence of ketone bodies and lactate as alternative fuels. Thus, it is now recognised in Europe and North America that, for this group, it is not appropriate to carry out routine blood glucose monitoring, to label low blood glucose as a pathological entity or to initiate treatment which is invasive or which may interfere with the establishment of breastfeeding ( ; ; ; ). Because of the healthy infant’s ability to counterregulate, problems with establishment of successful breastfeeding are equally likely to present with excessive weight loss (in excess of 10% birthweight), dehydration and jaundice as with clinically significant hypoglycaemia. Therefore, breastfeeding advice and intervention should not be based on blood glucose levels, but on full assessment of the baby – proceeding to blood glucose measurement if there are clinical concerns. Midwives and doctors must be alert to the possibility that other conditions, such as infection or, more rarely, inborn errors of metabolism, may present with the neurological signs and hypoglycaemia and specific investigations should be performed ( Table 34.4 ).

At-risk babies ( Table 34.2 )

For practical purposes, the following discussion focuses only on the infants who are at risk of the impaired metabolic adapatation and the neurological sequelae of hypoglycaemia ( Table 34.2 ). Early prevention of hypoglycaemia is optimal for these infants, so the first step in management must be to identify them. Although this is easy in some cases (such as the preterm baby), for others clinical observations are important (for example, to identify the wasted appearance of the growth-restricted neonate who may not necessarily have a low birthweight).

At-risk infants should have regular pre-feed blood glucose monitoring (at least 4–6-hourly initially). In addition, it is imperative that any infant with neurological signs, even if not in an at-risk group, should have urgent, accurate blood glucose measurement. The monitoring schedule for at-risk infants will vary according to local protocols, but we suggest that monitoring should be commenced before the second feed and that pre-feed monitoring be continued until the infant has had at least two satisfactory measurements. Monitoring should be recommenced if the infant’s clinical condition worsens or energy intake decreases. If monitoring is by reagent strip, low levels must be confirmed by accurate measurement (see above).

The importance of early milk feeding has been appreciated for many years ( ). Both breast and formula milks provide important gluconeogenic precursors and fatty acids for β oxidation. As they contain sources of energy other than carbohydrate they have a higher joule/ml content than 10% dextrose. In addition, enteral milk feeding stimulates the secretion of gut hormones, which may facilitate postnatal metabolic adaptation ( ). Therefore, all infants who are expected to tolerate enteral feeds should be fed with milk as soon as possible after birth, and at frequent intervals thereafter. Babies who are capable of sucking should be offered the breast at each feed (if this is the mother’s wish). If it is likely that babies will need complementary or supplementary feeds, maternal breast milk expression should be encouraged. The need for formula supplementation will vary between babies, will diminish with the successful establishment of breastfeeding, and will be guided by the availability of expressed breast milk, regular pre-feed blood glucose monitoring, the clinical condition of the baby and assessment of breastfeeding. In the breastfed baby, formula intake should be kept to the minimum necessary, so as to enhance breastfeeding and avoid suppression of normal metabolic adaptation ( ).

In the at-risk baby who is establishing oral feeds there is a potential nadir at which body stores are steadily reducing but milk feeds have not yet started to replenish these stores. Even if the baby is feeding well, this may not occur until at least 48 hours. For this reason, vulnerable babies should not be transferred to the community at less than 48 hours, and only when experienced staff are satisfied that feeding is effective.

When full enteral feeding is not anticipated, for example in the very preterm or sick infant, an intravenous glucose infusion should be commenced as soon as possible after birth. Usually 10% dextrose at 3 ml/kg/h (5 mg glucose/kg/min) is sufficient to prevent hypoglycaemia, but in some cases (such as hyperinsulinism) more is required. If the amount of glucose administered is limited by fluid restriction, more concentrated dextrose solutions may be required and central venous lines should be used, because these solutions are sclerotic to peripheral veins and cause tissue damage if they leak.

If low blood glucose levels persist or are associated with clinical signs in the milk-fed infant despite the above measures, it may be possible to increase further the volumes and/or frequencies of feeds. If this is not possible, or if the hypoglycaemia is resistant to this strategy, intravenous glucose will be required. If the infant is tolerating milk feeds these should be neither stopped nor reduced. The initial rate of 10% glucose infusion should be 3 ml/kg/h (5 mg/kg/min; Table 34.5 ), but adjusted according to frequent accurate blood glucose measurements. If the need for fluid restriction limits the amount of glucose that may be given, more concentrated solutions may need to be infused ( Table 34.5 ). If hypoglycaemia persists despite intravenous glucose, it is important to check the infusion site and the infusion apparatus to confirm glucose delivery. Leaking drips should be promptly resited. Boluses of concentrated glucose solution should be avoided because of the risk of rebound hypoglycaemia and cerebral oedema ( ); if boluses are required (for example, if there are neurological signs of hypoglycaemia), they should be of 10% dextrose (3–5 ml/kg), given slowly, and always followed by an infusion. All reductions in infusion rate should be gradual. In cases of hyperinsulinism, intramuscular glucagon will have a temporary glycaemic effect if there is delay in siting an intravenous infusion (see below).

Specific treatments

Clinical significance

It is of the utmost importance to remember that neonatal hyperglycaemia may be a sign of a serious underlying disorder, such as infection. However, despite reports of associations between hyperglycaemia and adverse outcomes in exteremely low-birthweight babies, it is still not known whether the high glucose concentrations themselves place the infant at further risk or whether the high blood glucose levels simply reflect the fragile and unstable condition of the babies most at risk of adverse outcome ( ). Unlike adults with insulin deficiency, hyperglycaemic neonates do not develop ketosis or metabolic acidosis ( ). There is a risk that glycosuria and osmotic diuresis may cause fluid and electrolyte imbalance with dehydration. These disturbances are more likely to be comorbidities rather than the result of hyperglycaemia as studies of large numbers of infants have reported that osmotic diuresis is not an invariable consequence of neonatal hyperglycaemia ( ; ; ; ). There is also concern that changes in blood osmolality and fluid shifts may result in cerebral damage. However, cerebral pathology and adverse neurodevelopmental outcome have never been demonstrated to occur as the direct result of hyperglycaemia, and it is thought that blood glucose levels above 20 mmol/l are required to exert significant osmolar effects ( ; ; ).

As described above, neonatal hyperglycaemia is clinically significant in that it may herald a serious underlying disorder. Once such disorders have been ruled out and treated, there is no evidence that self-limiting hyperglycaemia secondary to immaturity of glucoregulation or excessive glucose intakes and not associated with osmotic diuresis has adverse effects at blood glucose levels below 20 mmol/l.

Definition and diagnosis

There is no established definition of neonatal hyperglycaemia, but blood glucose levels above 7 mmol/l are usually considered to be above the normal range and the majority of units recently surveyed define hyperglycaemia as a blood glucose level above 10 mmol/l ( ). However, the upper ‘safe’ limit of blood glucose concentration in the neonate is likely to be above this level. As with hypoglycaemia, there is great variation in terms of the diagnosis and management of hyperglycaemia ( ).

Glucose reagent strips are more useful in the diagnosis of hyperglycaemia than for hypoglycaemia because the strips are more reliable at high blood glucose levels, and inaccuracies of 0.5–1.0 mmol/l are of less clinical relevance in the context of hyperglycaemia. It may also be useful to monitor urine for glycosuria, but it should be remembered that neonates, particularly those who are preterm, have a low renal threshold for glucose and fractional excretion of glucose varies widely, so that glycosuria may be present even in normoglycaemia and is independent of circulating blood glucose concentration ( ; ).

Prevalence

Without a clear definition of hyperglycaemia it is difficult to comment on its frequency. Studies of prevalence vary according to their subjects, with hyperglycaemia found most frequently in very low-birthweight and preterm infants ( ; ). Small-for-gestational-age infants who are preterm are more at risk for developing hyperglycaemia than hypoglycaemia when receiving standard intravenous infusions ( ). Reported prevalences vary from 29% to 86% in very low-birthweight neonates ( ; ).

Mechanisms and at-risk groups

The mechanisms underlying neonatal hyperglycaemia vary and, as with hypoglycaemia, are best understood with reference to the expected metabolic changes at birth. In contrast to hypoglycaemia resulting from a low glucose production rate or a high glucose uptake rate, hyperglycaemia may be the result of a high glucose production or infusion rate or a low glucose uptake rate.

Neonatal hyperglycaemia is usually secondary to a high glucose appearance rate, that is, failure of the baby to suppress endogenous glucose production even if glucose infusion rate is high ( ; ; ). To maintain control, the infant must be able to adapt to the exogenous administration of glucose by suppressing glucose production by the liver. The ability to glucoregulate in this way has been demonstrated in normoglycaemic neonates ( ). However, there is evidence from clinical and animal studies that some neonates do not suppress glucose production in response to glucose infusion and/or increased blood glucose levels ( ; ; ).

The inability to suppress gluconeogenesis may in turn be the result of disordered glucoregulatory hormone control. Although the glucoregulatory role of insulin in the neonate is unclear and may vary between infants, it has been suggested that hyperglycaemia results from decreased insulin secretion in immature subjects ( ; ). This is analogous to the adult insulin-dependent diabetic. Animal studies have also shown that, after chronic hyperglycaemia, the fetal pancreas cannot mount an insulin response to a further glucose surge ( ). This may be analogous to the condition in preterm babies receiving constant high-rate glucose infusions, whose pancreatic response to hyperglycaemia may be ‘exhausted’.

Alternatively, circulating insulin concentrations may be appropriate for the blood glucose concentration, but hyperglycaemia may result from end-organ insensitivity to insulin. This is analogous to type 2 diabetes, which is characterised by insulin resistance. Neonatal insulin resistance has been demonstrated by the persistence of hyperglycaemia in the presence of raised insulin concentrations, by the poor hypoglycaemic response to large exogenous doses of insulin and by the high insulin concentrations needed to suppress gluconeogenesis ( ; ; ; ). Insulin resistance may be secondary to immaturity or downregulation of peripheral receptors, to the effect of high fatty acid levels resulting from infusion of fat emulsion or to the peripheral actions of counterregulatory hormones ( ).

Clinical data demonstrate that some hyperglycaemic preterm infants have inappropriately low plasma insulin concentrations and high plasma catecholamine levels, whereas others have apparently appropriate insulin concentrations and may have insulin resistance ( Fig. 34.5 ) (Hawdon et al. unpublished data). The former group were those who had other clinical complications such as infection, and the latter were either very preterm or preterm and small for gestational age.

Plasma insulin concentrations in preterm infants who were clinically stable or who had clinical complications such as infection or necrotising enterocolitis (sick infants).

The excess secretion of counterregulatory hormones, which themselves stimulate glycogenolysis and gluconeogenesis, may in addition block the secretion of insulin and inhibit its peripheral action, thereby contributing to insulin resistance ( ). This is the mechanism of hyperglycaemia secondary to exogenous corticosteroids, sometimes administered in large doses to neonates with lung disease (The ). It has even been suggested that antenatal corticosteroid administration contributes to the failure to suppress postnatal gluconeogenesis ( ). Aminophylline, used for the prevention of apnoea of prematurity, mimics the action of catecholamines and induces glycogenolysis ( ). To date, growth hormone has not been implicated in the aetiology of neonatal hyperglycaemia ( ).

These hormonal disturbances may be the consequence of underlying clinical stresses such as infection, respiratory distress, pain or surgery ( ; ; ; ; ; ). Studies by ; ; ) demonstrated that, with minimal anaesthesia for major surgical procedures in term and preterm neonates, high glucagon and catecholamine levels and inhibition of insulin secretion led to a number of metabolic abnormalities, including hyperglycaemia and hyperlactataemia. This response was in proportion to the severity of surgical stress and could be prevented by the addition of opioid analgesia or halothane to anaesthetic regimens. In some conditions associated with severe clinical stress, such as perinatal asphyxia, hyperglycaemia may occur as the result of high counterregulatory hormone concentrations but hypoglycaemia is more often seen, probably because the latter represents the situation found after the stress response is exhausted.

Despite the frequency with which hyperglycaemia is now observed, the aetiology and optimal management of this metabolic disorder have not been established. Clinical, animal and laboratory studies suggest that glucose production is not suppressed in the face of hyperglycaemia, but there are conflicting data regarding the role of defective glucoregulatory hormone responses.

Prevention and management ( Fig. 34.6 )

As neonatal hyperglycaemia is usually self-limiting and not associated with adverse sequelae, many clinicians choose not to treat raised blood glucose concentrations aggressively. However, the first step in management, especially in a baby who has previously been normoglycaemic, must be to seek and treat serious underlying disorders. The second step is to prevent the occurrence of high blood glucose concentrations secondary to high glucose infusion rates by instituting careful management of intravenous fluid prescriptions. Clinicians often increase fluid infusion rates to counter renal and extrarenal losses in the immature neonate. The condition is usually self-limiting and may be prevented by the more gradual increase in glucose intake in those at risk. For example, 200 ml/kg/day of 10% dextrose provides 14 mg/kg/min glucose, which is well in excess of the neonate’s requirements. Thus it is not surprising that hyperglycaemia occurs and the extra glucose given is ‘wasted’. Therefore, glucose infusion rates should be calculated and, if they are found to be excessive (for example, above 4–6 mg/kg/min), more dilute solutions should be used.

Steps in the management of neonatal hyperglycaemia. BG, blood glucose.

Hyperglycaemia may still occur in some neonates who are clinically stable and who are not receiving excessive glucose intakes. These infants are usually of extremely low birthweight and less than 1 week old. Often they have received early parenteral nutrition and thus fairly high rates of glucose infusion in combination with amino acids. At the same time they may have high counterregulatory hormone levels, rendering them ‘catabolic’ with or without peripheral insulin resistance, so that they cannot utilise the infused substrates. The condition is usually self-limiting and may be prevented by the more gradual increase in glucose intake in those at risk.

There are three strategies for management of hyperglycaemia when it occurs in these circumstances.

First, moderate hyperglycaemia may be ‘tolerated’ if it does not appear to be causing osmotic diuresis. In most instances the condition resolves within a few days even if no action is taken.

Second, the rate of glucose infusion may be carefully reduced to the rate at which blood glucose levels become normal, and then gradually increased as tolerated. This carries the possible disadvantage of reducing the infant’s energy intake, but it is likely that immature infants are unable to utilise effectively all the glucose offered, especially if at a rate in excess of 5 mg/kg/min ( ).

Third, insulin may be administered in order to lower the blood glucose concentration without reducing the glucose infusion rate. If insulin is to be used, there needs to be priming of infusion tubing as insulin is adsorbed to plastic to a variable extent. Controlled studies of insulin administration to adult intensive care patients on intravenous nutrition have demonstrated that, although there is a short-term improvement in nitrogen balance, there is no advantage in terms of either weight gain or body composition, and that a number of patients become hypoglycaemic ( ). Although there are a number of reports of this practice in the neonatal literature, there is no consistency regarding the clinical situations in which insulin has been given, only short-term outcome measures have been reported and there are few prospective controlled trials ( ; ; ; ; ). All of these studies reported hypoglycaemia in some infants even after the discontinuation of insulin infusion. Therefore, insulin should not be prescribed without availability of prompt and accurate blood glucose testing. Recent systematic reviews have not supported the strategy of routinely infusing insulin to achieve a specific glucose delivery rate or prevent neonatal hyperglycaemia, and recommend reserving the use of exogenous insulin for infants with severe hyperglycaemia (above 14–16 mmol/l) because the safety of the practice has not yet been determined ( ; ).

There is marked variation between and within studies regarding the doses of insulin required, and some authors have raised the possibility of the development of tolerance in some babies, so that hyperglycaemia recurs despite large increases in insulin dosage ( ; ; ). Considering the clinical observation that without insulin treatment neonatal hyperglycaemia is usually a transient and self-limiting condition, it may be that insulin administration in some way hinders spontaneous recovery.

Epidemiological studies have suggested that intrauterine nutritional and endocrine status may influence adult metabolism and susceptibility to disease ( ). The long-term clinical significance of early high energy intakes in association with large doses of exogenous insulin in the preterm neonate, and of possible up- or downregulation of insulin receptors, has not yet been considered.

Finally, the introduction of enteral feeding with small volumes of milk as soon as the infant’s gastrointestinal tract will tolerate this may hasten the control of blood glucose homeostasis by inducing surges of gut hormones which promote insulin secretion (the enteroinsular axis) ( ).

Accidental hyperglycaemia should be prevented by strict attention to stocked glucose solutions, pharmacy total parenteral nutrition production quality control and glucose infusion devices.

The pathogenesis of neonatal hyperglycaemia and the mechanism of action of exogenous insulin administration are not clearly understood. Therefore, the increasing practice of prescribing insulin to neonates without understanding its mode of action is of concern. It is not known whether insulin administration promotes linear growth in neonates, or merely converts glucose into fat. Further prospective randomised controlled trials of insulin therapy in preterm neonates who become glucose-intolerant would assess the impact of insulin therapy on glucose tolerance, metabolism and growth in the short and long term. It is currently impossible to judge whether insulin therapy does confer clinical advantage over expectant management of a condition which is usually self-limiting.

Molecular basis

The molecular background to some cases of PHD has been established. Isolated deficiencies of GH ( ; ), TSH ( ), ACTH ( ; ) or gonadotrophins ( ; ; ; ) may reflect mutations of the respective genes. Mutations of the GH-1 gene can result in recessive and dominantly inherited isolated GH deficiency ( ; ; ; ) and defects of the GHRH receptor can also result in isolated GH deficiency ( ). Severe isolated ACTH deficiency presenting in the neonatal period may be due to mutations in TBX19 ( TPIT ) ( ).

A defect in one of the genes involved in the development of the pituitary may result in a more severe endocrinopathy with deficiencies of more than one hypothalamic and/or pituitary hormone. Defects of the gene POU1F1 ( Pit-1 ) are typically associated with a deficiency of GH, TSH and prolactin ( ; ; ; ) whilst the gene PROP1 , active at an earlier stage of pituitary gland development, is associated with deficiency of GH, TSH, prolactin, gonadotrophins ( ; ) and sometimes ACTH deficiency as well ( ; ) ( Table 34.6 ). HESX1 has a fundamental role in CNS development and homozygous and heterozygous mutations of HESX1 are a recognised cause of SOD. Heterozygous defects can also cause isolated GHD ( ; ; ). Defects of SOX3 may result in PHD in association with mental retardation whilst defects in LHX4 may result in PHD in association with abnormalities of the cerebellar tonsils.

Clinical features ( Table 34.7 )

Clinical recognition of hypopituitarism in the newborn can be difficult. GH deficiency leads to relatively subtle changes in body form and whilst some authors report a reduction in length of about 1 sd with a normal birthweight ( ) other studies indicate that birth length is largely unaffected ( ; ). GH-deficient children grow slowly in infancy and may be more than 2 sd below the mean for length by the end of the first year of life ( ). Micropenis can be a feature of GH deficiency but is more typically associated with gonadotrophin deficiency ( ). GH promotes longitudinal growth but also has lipolytic and ketogenic activity. Ketones are an important fuel in the newborn period and babies with isolated GH deficiency are therefore more susceptible to hypoglycaemia. Hypoglycaemia is typically more profound in the infant with adrenocortical insufficiency and a deficiency of one or both of these hormones should be considered in the hypoglycaemic baby. Prolonged neonatal jaundice with a conjugated hyperbilirubinaemia is a further well-recognised feature of congenital hypopituitarism ( ) and is closely linked to cortisol deficiency. All neonates with suspected PHD should have an opthalmological assessment because of the possiblility of underlying SOD.

Further investigations ( Table 34.7 )

Diagnosing PHD typically involves piecing together a range of clinical and biochemical parameters. The neonate with microphallus, hypoglycaemia and a low free T ₄ (fT ₄ ) is at high risk of having CPHD. The diagnosis is even more likely if the infant has a conjugated hyperbilirubinaemia and small optic discs on opthalmological examination. Random measurements of pituitary hormones (particularly TSH and fT ₄ ) can be helpful but hypoglycaemia may not be an effective stimulus for GH and cortisol production (babies who prove to be GH- and ACTH-‘sufficient’ may have low levels at the time of hypoglycaemia). Hence stimulation tests may be required as part of the diagnostic work-up. Biochemical testing can be extremely helpful but requires careful planning and is potentially misleading because, like all tests, there are false positives and negatives. GH levels are often elevated in the first days of life and a low random GH concentration raises the possiblity of hypothalamopituitary disease.

It is important to remember the TSH surge and associated increase in thyroid hormone levels that occur in healthy babies around the time of delivery when interpreting thyroid function tests ( Fig. 34.8 ). Laboratories may not have appropriate age-related data for babies in the first days of life and so a ‘normal’ adult fT ₄ may be abnormally low at this time of life. TSH levels can, paradoxically, be mildly elevated in secondary hypothyroidism and there may be an exaggerated response to TRH ( ). The increase in TSH levels and exaggerated response in hypothalamic hypothyroidism may reflect the presence of detectable but bioinactive TSH ( ) or alternatively the abnormal production of TSH in the absence of hypothalamic hormones such as somatostatin. Although the TRH test has its advocates ( ), many feel that the diagnosis of secondary hypothyroidism can be made with a single or serial measurements of TSH and thyroid hormone ( ).

Cortisol deficiency secondary to ACTH deficiency can be associated with hyponatraemia because cortisol inhibits arginine vasopressin (AVP) production and is required to excrete a water load. Potassium levels are not typically raised because mineralocorticoid production is primarily regulated by the renin–angiotensin system and not ACTH. The Synacthen test is still an appropriate test with which to investigate possible secondary hypoadrenalism (ACTH deficiency) prior to commencing hydrocortisone in the infant with hypoglycaemia and suspected hypopituitarism. The CRH test has theoretical advantages but both tests are hampered by concerns about sensitivity and specificity, a lack of reliable normative data as well as the more complex and, in the case of the CRH test, more limited availability of the ACTH assay. GnRH administration with assessment of LH and FSH levels may be helpful ( ) but the assessment of other axes is more important if there is a limited amount of blood available. It should be remembered that drugs such as dopamine may interfere with endocrine investigations ( ).

It can be difficult to balance the desire to obtain comprehensive biochemical data with the need to treat the sick, hypoglycaemic neonate. If the clinical picture strongly suggests hypopituitarism then it is wise to treat and then reassess the neonate biochemically at a later stage. Hormone deficiencies can evolve with time (see above) and a neonate who is ACTH-‘sufficient’ can become ACTH- and hence cortisol-‘insufficient’ in later childhood.

MRI can provide extremely helpful information about pituitary/hypothalamic anatomy as well as information about other CNS developmental abnormalities such as optic nerve hypoplasia or migrational abnormalities such as schizencephaly, which are more common in babies with CPHD ( Fig. 34.7 ).

Treatment

Hypopituitarism is treated by replacing the missing pituitary hormone (GH) or the product of its target organs (T ₄ or hydrocortisone). Although the cortisol production rate in the newborn is 6.6–8.8 mg/m ² /24 h ( ), the absorption and bioavailability of hydrocortisone can vary ( ). Adrenal replacement (secondary to ACTH deficiency) should be with hydrocortisone in a dose around 8–10 mg/m ² /24 h (in three or four divided doses). There are concerns about the bioavailability of some hydrocortisone suspensions and these need to be borne in mind ( ). For thyroid replacement, T ₄ in a single dose of approximately 8–10 µg/kg/24 h in the newborn is suitable initially. The objective is to obtain T ₄ levels in the upper part of the normal range and the dose should be adjusted accordingly. TSH levels will be low in secondary hypothyroidism once therapy has commenced and cannot be used to guide replacement. GH replacement may be needed to control the hypoglycaemia of hypopituitarism if this proves intractable and therapy may also be warranted from an early stage if growth is to be optimised.

Treatment

Treatment is usually with the vasopressin analogue DDAVP. This compound is more potent than the native molecule and has a longer half-life ( ). It is important to be aware of the danger of water overload, which will develop if urine is inappropriately concentrated for a given fluid intake. Babies have been treated with DDAVP administered nasally but the likelihood of water intoxication is markedly reduced with oral therapy and this is the preferred route of administration ( ; ). Capillary or serum sodium osmolality should be checked regularly, particularly in the initial phase as the dose is being adjusted. It is safer in both the short and the long term for an infant with diabetes insipidus (and an intact sense of thirst) to be allowed to compensate for a relatively low dose of DDAVP by feeding or drinking more and babies should be offered water on occasions as well as milk. Parents should be actively involved in management from an early stage and encouraged to gauge fluid input and output.

A suitable starting dose of oral DDAVP is around 1–4 µg daily, increased to twice a day once the impact of the first dose has been established. The dose can be administered orally using the desmopressin injection solution (4 µg/ml), although some people have used an aliquot of a DDAVP tablet dispersed in water. If the baby is sleeping longer overnight then the logical time to administer a single daily dose or the larger dose of a twice-daily regimen is in the evening. We do not recommend the nasal route of administration but if there is no alternative then the standard solution (100 µg/ml) can be diluted with 0.9% NaCl solution to create a more manageable volume (for example, 10 µg/ml – discuss with the hospital pharmacy). This can then be instilled into the nose using a 1-ml syringe. A suitable starting dose in the newborn is 0.1–1.0 µg once or twice daily. Initially it is wise to administer the second dose only when the impact of the first has been established.

Low-dose subcutaneous DDAVP has been advocated as a safe means of managing CDI in infancy ( ). A low dose (usually 0.01 µg, s.c.) used once a day can be initiated and then increased to maintain a normal serum sodium. DDAVP (4 µg/ml) can be diluted 1 : 10 in a syringe with the required dose ranging from 0.02 to 0.08 µg DDAVP s.c. twice daily.

Finally, 5 mg/kg of chlorothiazide administered twice a day has also been used to treat CDI on the basis that, as with nephrogenic diabetes insipidus, the associated increase in urine osmolality will compensate for vasopressin deficiency ( ).

Thyroid function in utero

TSH, modulated by TRH, stimulates the synthesis and release of T ₄ and triiodothyronine (T ₃ ) from the thyroid into the plasma, where they circulate, strongly bound to proteins. Maternal T ₄ , but not TSH or T ₃ , can cross the placenta in physiologically significant amounts and this explains the relatively normal phenotype in infants with CHT at birth with levels that are one-third to one-half normal values in babies unable to manufacture thyroid hormone ( ). The importance of maternal T ₄ delivery to the fetus in early gestation, whilst the fetal thyroid axis is developing, is supported by studies of neurodevelopment in infants of mothers with relatively low T ₄ levels ( ; ). The significance of T ₄ transfer in the latter stages of pregnancy is illustrated by the life-threatening illness seen in hypothyroid infants of hypothyroid mothers ( ; ).

The greater part of circulating T ₃ is derived from peripheral monodeiodination of the outer ring of T ₄ , which therefore acts as a reservoir or prohormone for the more active T ₃ . An alternative monodeiodination affects the inner ring of T ₄ and produces inactive reverse T ₃ (rT ₃ ). This mechanism permits a balance between production of the most and least active thyroid hormones. There are three deiodinases – type I (D1), which has inner and outer deiodination activity; type II deiodinase (D2), which catalyses outer-ring deiodination; and type III deiodinase (D3), which catalyses inner-ring deiodination.

Fetal thyroid hormone metabolism is characterised by a predominance of D3 activity outside the CNS in tissues such as the liver, kidney and placenta. This converts maternal and endogenous T ₄ preferentially to rT ₃ and presumably helps to reduce tissue thermogenesis and enhance tissue anabolism. T ₄ (rather than T ₃ ) is required by the developing brain, and the appropriate deiodinases (particularly D2) are expressed in a temporal and spatial manner in different brain regions. Sulphotransferase enzymes also have an important role in thyroid hormone metabolism. T ₄ sulphation blocks outer-ring deiodination to T ₃ whilst promoting inner-ring conversion to inactive rT ₃ . The activity of sulphotransferase enzymes in tissues like the liver will also, therefore, play a key role in determining thyroid hormone availability ( ). A key determinant of thyroid hormone transport into cells has been identified with defects of this protein resulting in a phenotype that includes profound neurodevelopmental delay ( ).

Thyroid function in the neonatal period

After birth there is an acute discharge of TSH which reaches a peak at 30 minutes, before falling towards basal levels within the first 3 days. There is an associated release of thyroid hormones and enhanced peripheral conversion of T ₄ (closely linked to D2 activity), which results in a pronounced increase in T ₃ in the first hours of life. There is a further increase in total T ₃ and free T ₃ levels for about 36 hours around the time of the postnatal peak in T ₄ ( Fig. 34.8 ) and fT ₄ levels remain relatively high for the first weeks of life ( ). In preterm infants the levels of T ₃ remain higher than the cord values of babies of equivalent gestational age for several weeks ( ). Preterm infants (down to 30 weeks) show similar but lesser changes in TSH and thyroid hormone concentrations, although the most preterm babies (23–27 weeks) have relatively low T ₄ values ( ). By 1–2 months of age, thyroid hormone levels are comparable to those in term infants. In the case of infants under 30 weeks’ gestation, the postnatal surge does not occur ( ) and T ₄ levels frequently fall to a nadir around 1–2 weeks of age, which is more pronounced with increasing prematurity ( ; ). T ₄ levels remain below those of full-term infants through the first few weeks of life ( ) and climb gradually to normal postnatal levels.

The trend in fetal and neonatal plasma thyroid-stimulating hormone (TSH), thyroxine (T ₄ ), triiodothyronine (T3) and reverse T3 (rT3) levels.

(Adapted from .)

Thyroid dysgenesis

Embyrological defects resulting in thyroid dysgenesis are the most common cause of CHT, and range from thyroid agenesis (no gland present) to an ectopic and/or hypoplastic gland. Some thyroid tissue is demonstrable in approximately two-thirds of affected infants, and there is a spectrum of biochemical severity. The frequency is around 1 in 3000–4000 and is twice as common in females as in males, with only minor racial and seasonal differences in incidence described. As genetic factors involved in thyroid development and thyroid-specific gene expression have been described, so abnormalities of these same genes have been identified in a small number of patients with thyroid dysgenesis. Mutations in the TSH receptor (TSHR) ( ; ) and transcription factors such as TTF1 (NKX2-1), TTF2 and Pax-8 have been published, but a strict mendelian inheritance pattern is relatively uncommon ( ; ) and the aetiology of thyroid dysgenesis in most patients is unknown ( ).

Thyroid hormone replacement abolishes the physical manifestations of hypothyroidism but intellectual and neurological prognosis is compromised in those babies at the more severe end of the spectrum unless treatment is started within the first few weeks of life ( ).

Dyshormonogenesis ( Fig. 34.9 )

Recessively inherited biochemical defects of iodothyronine synthesis are the second most common cause of permanent CHT, with a frequency in Europe and North America of approximately 1 in 30 000–50 000, accounting for some 10–15% of patients identified by screening. The sex incidence is equal. A goitre may be present at birth but, unless large, may be difficult to detect and may not develop until later in life. The degree of hypothyroidism is variable with some patients clinically and even biochemically euthyroid. Dyshormonogenesis can be a cause of transient hypothyroidism when the extra demands in early life cannot be met. Thyroid ultrasonography usually shows normal thyroid anatomy and site (a eutopic gland). In many dyshormonogeneses there is excessive discharge of iodide from the thyroid gland after administration of perchlorate (the perchlorate discharge test), due to an abnormality of iodide organification into thyroglobulin. Recognised abnormalities of thyroid hormone synthesis include defects in iodide transport across the basolateral membrane (sodium iodide symporter) ( ; ) and then into the colloid lumen ( ; ). Other causes include abnormalities of the thyroglobulin gene ( ), thyroid peroxidase ( ; ) and THOX ( ), which is involved in the generation of H ₂ O ₂ required by thyroid peroxidase to oxidise iodide. Delineation of the precise biochemical defect in these disorders does not alter the need for T ₄ replacement, but can help when counselling families about the likelihood of recurrence.

Dyshormonogenesis. Thyroid hormone production and disorders of thyroid hormone synthesis (dyshormonogeneses). Thyroid hormone production: amino acids and glucose are used to produce thyroglobulin to which tyrosine residues are attached. Iodine is incorporated into these residues at the apical membrane to produce iodotyrosines. These are stored as extracellular colloid. Thyroid hormone secretion: colloid droplets are invaginated to release thyroglobulin and the iodotyrosines, including thyroxine (T ₄ ) and triiodothyronine (T ₃ ), which are released into the circulation. The other iodotyrosinases (DIT and MIT) are deiodinated and the iodine reused. Dyshormongeneses: a, thyroid-stimulating hormone (TSH) receptor defects; b, abnormalities of iodide uptake (sodium iodide symporter); c, defects of thyroglobulin synthesis; d, abnormalities of iodine transport at the apical membrane (pendrin); e, abnormalities of the enzyme system incorporating iodine into tyrosine (thyroid peroxidase); f, abnormalities of the enzyme system involved in H ₂ O ₂ generation (Thox); g, abnormalities of the deioidinase system.

Diagnosis

The diagnosis of CHT is confirmed by finding low serum thyroid hormone concentrations (fT ₄ ) and a high TSH. The fT ₃ level is variable and is preserved in the normal range for longer. It remains essential to exclude hypothyroidism in any infant in whom there is clinical suspicion, because, first, errors inevitably occur in all screening programmes; second, mild hypothyroidism may escape detection by all screening methods (see below); and third, there may be a late rise in TSH levels. The major clinical features of CHT are summarised in Table 34.10 . Infants with low T ₄ values are more likely to have clinical signs such as feeding difficulties, lethargy, prolonged jaundice, umbilical hernia and macroglossia ( ). Such infants are more likely to have an aplastic rather than a hypoplastic or ectopic gland. A gestation over 40 weeks, induction of labour and a birthweight above 3500 g are all more common among infants with hypothyroidism.

Further investigation

The laboratory tests of thyroid function primarily determine whether an infant requires T ₄ treatment. Isotope scanning (technetium or ¹²³ I) can be undertaken before or within a few days of starting T ₄ replacement treatment, ideally when the TSH levels are still high ( ). Thyroid ultrasonography is also performed in some centres ( ) and, whilst it is a highly operator-dependent technique, it can provide information about gland morphology and hence more detail than isotope scanning alone. An ultrasound can be conducted on the neonatal unit and may be of particular value in the infant exposed to iodine-containing compounds whose isotope scan may be hard to interpret.

In thyroid dysgenesis, an isotope scan or ultrasound will reveal an absent, small or ectopic gland. This condition has a low risk of recurrence in further children ( ). The infant with thyroid aplasia or ectopia will need treatment for life and is unlikely to benefit from a trial off therapy in later childhood. If the gland is anatomically normal and the isotope uptake is high, then hypothyroidism is probably due to a dyshormonogenesis and may be inherited in an autosomal recessive (AR) manner. Absent isotope uptake in a normally sited gland on ultrasonography also suggests a dyshormonogenesis with abnormal iodine trapping.

Advocates of investigations in CHT highlight the fact that this information can be used to refine the T ₄ dose ( ) and can provide information about recurrence risk and prognosis. On the other hand investigations can be operator-dependent, difficult to interpret and may involve intravenous access and ionising radiation. Some clinicians still prefer to rely on the biochemistry (TSH and thyroid hormone levels) alone. A plain radiograph of the knee to assess epiphyseal maturity was frequently performed in the past. This may reflect the degree of fetal hypothyroidism but is of limited clinical value.

A TRH stimulation test is rarely indicated, but has been used in babies with mild thyroid dysfunction. An exaggerated and prolonged TSH response is observed if TSH levels are mildly elevated because of a compromised thyroid gland ( ). TRH testing has been advocated as a way of teasing out central hypothyroidism in babies found to have low T ₄ levels on screening ( ).

Treatment

T ₄ is used for replacement and there is no advantage in giving T ₃ (see above). It is a matter of some urgency to start treatment, and when the diagnosis is reasonably certain it may be preferable to collect further specimens and start treatment without waiting for the results.

The aim of therapy is to normalise T ₄ and TSH concentrations quickly, and an appropriate starting dose of T ₄ in the newborn with low T ₄ levels is 10–12 µg/kg/24 h given as a single daily dose ( ; ; ). Full-term infants of normal birthweight should be commenced on 37.5–50 µg daily. This will usually bring the TSH into the normal range quickly and the dose of T ₄ may then need to be reduced. Babies with significant endogenous thyroid hormone production may normalise TSH values on a smaller dose. The smallest tablet available contains 25 µg T ₄ but as the hormone has a long half-life (in excess of a week) the dose can be adjusted in smaller increments by (for example) giving a higher or lower dose on alternate days as well as by dividing them. T ₄ solutions are available, although there have been concerns about the stability of some of these preparations and parents need to be aware that they may come in different strengths.

Relatively high fT ₄ levels may be seen after treatment is commenced but this picture does not necessarily indicate overtreatment if TSH levels are normal or raised. The dose of T ₄ will fall with time when calculated according to weight and a suppressed TSH indicates overtreatment. The heterogeneity of CHT has been highlighted earlier and babies with only mild or moderately elevated baseline TSH levels may require smaller doses of T ₄ than babies with agenesis, in whom the baseline TSH is typically very high ( ). Replacement is usually satisfactory if the serum total T ₄ and fT ₄ are in the upper part of the quoted laboratory normal range for age, but suboptimal compliance needs to be excluded if TSH values remain elevated despite a substantial T ₄ dose.

There is evidence to suggest that higher initial starting doses of T ₄ may have a beneficial impact on neurodevelopment ( ; ; ), although there has been concern that this might be at the expense of an increase in behavioural problems ( ; ). Severe overdosage may also cause accelerated growth and even craniosynostosis, but the danger of impaired neurological development from underdosage is probably greater.

Higher doses of T ₄ treatment in early life as well as a milder baseline biochemical picture have been associated with increased stature in mid-childhood and at final height ( ; ; ), although this has not been a consistent observation ( ).

Follow-up and prognosis

Growth, clinical progress and thyroid function should be checked after 2 and then 4 weeks on treatment, and then at 1–2-monthly intervals in the first 6 months and then every 3–4 months up to 3 years of age. Definitive reassessment of thyroid status can be undertaken after the age of 2 years, when brief cessation of treatment will have no adverse effects. Most children with a permanent defect of thyroid hormone generation will develop abnormal biochemistry without any adverse symptoms if T ₄ is stopped for 2–4 weeks at this stage. In addition to a biochemical assessment of thyroid function, the opportunity can be taken to obtain or repeat an isotope thyroid scan. If the T ₄ dose has been increased during the first 2 years of life because of an elevated TSH then this may not be necessary.

CHT is associated with impaired motor development, a reduced intelligence quotient (IQ) and impaired hearing and language problems ( ; ; ). These features are related to the severity of hypothyroidism at birth ( ). In one study, total T ₄ levels below 42 nmol/l in the neonatal period were associated with a 10-point deficit in IQ at school entry ( ). Individuals with T ₄ values above 42 nmol/l were no different from controls, suggesting a ‘threshold’ effect. Thus it seems that the degree of pre- and perinatal hypothyroidism has implications for CNS function in the long term. A formal hearing assessment is recommended in the hypothyroid infant because of the 10-fold increase in hearing loss in these babies ( ).

Transient hypothyroidism and hyperthyrotropinaemia (subclinical hypothyroidism)

In this heterogeneous group of disorders, the TSH is raised and thyroid hormone concentrations are within the age-related reference range (hyperthyrotropinaemia) or temporarily low (hypothyroidism). Babies may have had a mild abnormality of thyroid function detected by the screening programme or may have had thyroid function checked for other reasons. Although thyroid function may normalise within the first few months of life, some of these infants may have a permanent, albeit subtle, abnormality of thyroid function ( ; ). There are well-established causes of transient thyroid dysfunction but some of these children will lie at the mild end of the CHT phenotypic spectrum. The overlap with CHT is illustrated by the transient hyperthyrotropinaemia that can occur in infants with an underlying dyshormonogenesis ( ). These babies are presumably unable to meet the demand for T ₄ in the early neonatal period. Some babies with hyperthyrotropinaemia will have abnormal thyroid gland morphology ( ).

A list of causes of hyperthyrotropinaemia/transient hypothyroidism is given in Table 34.11 . Noteworthy causes of transient hypothyroidism include intrauterine exposure to antithyroid drugs ( ), iodine deficiency ( ; ) and exposure to topical iodinated antiseptic agents. The last are readily absorbed and should be avoided or carefully removed following initial application ( ; ; ). Transient hypothyroidism may also be due to the transplacental passage of thyrotropin receptor-blocking antibodies in approximately 2% of cases of CHT ( ) and it is wise to check a mother’s thyroid status in these circumstances. Causes of permanent hyperthyrotropinaemia include abnormalities of thyroid gland development ( ) and abnormalities of the TSH receptor ( ). Abnormal thyroid imaging is, not surprisingly, more common in babies with permanent hyperthyrotropinaemia ( ). Transient hypothyroidism and hyperthyrotropinaemia appear to be more common in the preterm infant with some infants demonstrating the delayed TSH rise described above. Hence iodine deficiency or exposure to excess iodine could result in a ‘delayed’ TSH rise as well as transient hypothyroidism and hyperthyrotropinaemia.

Intervention is important in those patients with low thyroid hormone levels and some clinicians may treat relatively subtle increases in TSH that persist on the basis that this is a marker of thyroid hormone insufficiency within the developing CNS ( ; ). Intervention is of unproven benefit but likely to be safe provided the infant is monitored appropriately and TSH suppression avoided.

Transient hypothyroxinaemia ( )

All premature infants have some degree of hypothyroxinaemia because cord serum T ₄ values increase with gestational age. Low T ₄ levels in the preterm infant reflect the loss of the normal maternal T ₄ supply, immaturity of the hypothalamopituitary–thyroid axis, as well as other factors such as the ongoing fetal tendency to convert thyroid hormone to inactive rT ₃ . Low thyroid hormone values persist in the first 1–2 weeks in association with low fT ₄ and normal TSH levels. This biochemical picture tends to be more profound in the most preterm babies ( ). This was thought to be of little consequence, but more recent evidence indicates that there is a relationship between T ₄ concentrations in preterm and low-birthweight (LBW) infants and subsequent neurodevelopmental outcome ( ; ; ; ).

The impact of thyroid hormone administration on the short- and longer-term outcome following preterm delivery has yet to be established. Improvements in survival have been reported in some studies ( ) but not others ( ) and early neurodevelopmental outcome does not appear to be affected by treatment. A subgroup analysis in the study by ) suggested a potential benefit of T ₄ treatment (8 µg/kg/day) on the Bayley Mental Development Index in babies of 25–26 weeks’ gestation. However, evidence from the same and other studies has suggested that increasing T ₄ delivery in babies born at 27–29 weeks may have an adverse effect on neurodevelopment. A more recent report from the same group, as the children reach school age, has confirmed the improved IQ with a reduction in behavioural problems in those receiving T ₄ at less than 27 weeks, and lower IQ with more developmental problems in those receiving T ₄ at 29 weeks ( ). The effects of T ₄ administration may therefore depend on the gestation and age of the infant treated, with protective mechanisms ensuring an adequate supply of T ₃ in all but the most preterm. At this time, the routine supplementation of babies of any gestation in whom TSH levels are normal is not recommended, but we do recommend following the thyroid hormone levels until they normalise.

It is important to remember that preterm infants with hypothyroxinaemia may have a permanent abnormality of thyroid function which will only become apparent with longer term follow-up ( ).

Secondary (hypopituitary) hypothyroidism

Secondary hypothyroidism (TSH deficiency) should be suspected if fT ₄ and T ₃ levels are low in the presence of a low, normal or paradoxically mildly elevated TSH. The differential diagnosis of this biochemical picture includes euthyroid illness and prematurity (see below). Isolated TSH deficiency is very rare and most infants with secondary hypothyroidism will have other PHDs with associated clinical and biochemical features. Screening programmes using the combination of TSH and T ₄ suggest that the incidence of secondary hypothyroidism is around 1 in 16 000. The TSH response to TRH in patients with secondary hypothyroidism may be normal, poor or even exaggerated ( ).

The ‘low T ₃ syndrome’ (non-thyroidal illness)

This term is used to describe thyroid function tests characterised by low serum T ₃ concentrations, normal or raised rT ₃ , variable T ₄ and normal TSH. Fetal T ₃ levels are low throughout gestation because of enhanced conversion of T ₄ to rT ₃ , and this picture is frequently observed in preterm infants. As in older patients, T ₃ levels may be further reduced by intercurrent illness and by poor nutrition in infants of all gestational ages. Low T ₃ levels, like T ₄ , have been linked to a reduction in IQ in later life ( ). T ₄ administration to infants of less than 30 weeks’ gestation does not increase T ₃ levels ( ).

Abnormalities of binding proteins: thyroxine-binding globulin deficiency and familial dysalbuminaemic hyperthyroxinaemia

The major carrier protein of the thyroid hormones is thyroxine-binding globulin (TBG). Deficiency of TBG is usually inherited as an X-linked dominant trait and has an incidence in male infants of 1 : 2400 with an overall frequency around 1:4000–4300 in North America ( ; ). Affected patients are euthyroid. Total T ₄ and T ₃ levels are low, but the free fractions are normal and the resting and stimulated TSH values are also normal. TBG measurement is needed for a definitive diagnosis.

Familial dysalbuminaemic hyperthyroxinaemia is a relatively common disorder, affecting 1 in 100 people. It is characterised by increased T ₄ levels in clinically euthyroid individuals because of an abnormal albumin molecule that has an increased affinity for thyroid hormone. The abnormal albumin molecule will also interfere with some thyroid hormone assays, leading to spurious and potentially confusing results ( ). This condition underlines the importance of measuring TSH concentrations in infants with raised thyroid hormone concentrations.

Abnormal thyroid hormone transport into the cell

Thyroid hormone transport across the cell membrane is an active process that involves the membrane protein monocarboxylase transporter 8 (MCT8). MCT8 action is crucial if thyroid hormone is to enter the cell and then exert its biological effects by binding to nuclear thyroid receptors. Mutations in MCT8, located on the X chromosome, result in severe psychomotor retardation with a biochemical picture that classically involves elevated T ₃ levels and a normal or mildly elevated TSH. Hemizygous males with MCT8 mutations are severely affected whilst heterozygous females with one abnormal and one normal allele have a normal phenotype in the presence of a milder biochemical abnormality ( ).

Thyroid hormone resistance ( ; )

Resistance to thyroid hormone is caused by mutations in the thyroid hormone receptor (TR beta) gene. The clinical presentation of resistance to thyroid hormone is highly variable and the majority of individuals are completely asymptomatic. Occasionally neonates will be found to have this disorder because of a goitre or because of a raised TSH. A minority may have symptoms suggestive of hypothyroidism such as growth retardation and impaired cognitive ability, whilst others have signs of thyroid hormone excess such as tachycardia or hyperactivity ( ). The typical picture is of a healthy, clinically euthyroid infant with a raised TSH and raised fT ₄ and fT ₃ . The disorder is usually inherited in an autosomal dominant (AD) manner and so a parent may have similar biochemistry.

Neonatal thyrotoxicosis

Neonatal thyrotoxicosis is a rare but potentially serious condition, usually caused by the transplacental passage of TSH receptor-stimulating antibodies from the serum of a mother with active, inactive or treated Graves disease ( ). It may also occur when the mother has autoimmune thyroid disease other than Graves ( ). The disease is usually transient, resolving within the timespan of the circulating antibodies. Neonatal hyperthyroidism is underrecognised but it can often be predicted and treated prenatally.

TSH receptor-stimulating antibodies may be demonstrated in the majority of patients with Graves disease and they cross the placenta freely. Thyrotoxicosis in the fetus may lead to preterm labour, LBW, stillbirth and neonatal death. Approximately 1% of babies of mothers with Graves disease become overtly thyrotoxic, although the absolute risk is related to the concentration of TSH receptor-stimulating antibodies in the maternal serum ( ; ; ) and some babies will be biochemically toxic but asymptomatic. Thyroid dysfunction has been detected in as many as 16.5% of babies born to women with Graves disease ( ). TSH receptor-blocking antibodies may be present as well, and the clinical and biochemical picture will reflect a range of factors, including the impact of altered maternal and fetal thyroid hormone levels on fetal hypothalamopituitary function as well as the nature and concentration of prevailing antibody concentrations. Exposure to antithyroid drug treatment in utero will also influence the neonatal picture. Hence, an infant who is initially hyperthyroid can subsequently become hypothyroid (requiring T ₄ replacement) and an initially hypothyroid infant can become hyperthyroid ( ; ; ).

Although neonatal thyrotoxicosis is usually due to the transplacental passage of TSH receptor-stimulating antibodies, activating mutations of the TSH receptor can also give rise to a similar picture. This form of hyperthyroidism is typically inherited in an AD fashion (a parent may have a history of thyroidectomy) and will not resolve spontaneously. Treatment is with total thyroidectomy ( ; ).

Clinical features ( Table 34.12 )

The signs of thyrotoxicosis in the fetus include tachycardia and intrauterine growth retardation. Infants with perinatal thyrotoxicosis may show signs of hyperthyroidism immediately after birth, but symptoms may be delayed as long as 4–7 weeks ( ; ). This may be due to the effect of maternal antithyroid drugs or to the relative effects of both blocking and stimulating antibodies. Infants may have a palpable goitre, and although eye signs, especially proptosis and lid retraction, can be present at birth, they are often mild or absent throughout the course of the disease. Rarely, a mother with euthyroid ophthalmic Graves disease may produce an infant with eye involvement but no evidence of thyrotoxicosis. Signs of CNS stimulation, such as irritability, restlessness and jitteriness, usually predominate and there is tachycardia and occasionally arrhythmia, which may progress rapidly to severe and intractable heart failure. Other signs of hypermetabolism include an excessive appetite with weight loss or inadequate weight gain, diarrhoea, sweating and flushing. Less predictable clinical features include hepatosplenomegaly, jaundice and accelerated bone maturation, which can cause premature closure of the skull sutures. Mortality rates of 16–25% have been reported and the long-term outcome is uncertain ( ).

Management ( Fig. 34.10 )

Hyperthyroid pregnant women should be treated with the antithyroid thiourea derivatives carbimazole or propylthiouracil ( ). The use of radioactive iodine is absolutely contraindicated during pregnancy ( ) and surgery may precipitate preterm delivery. The antithyroid drugs cross the placenta, and the lowest dose that controls the hyperthyroidism should be used. Propylthiouracil is excreted into breast milk in lower concentrations than carbimazole and hence may be preferable. The ‘block and replace’ regimen, using higher doses of antithyroid drugs in combination with replacement T ₄ , should be avoided. If maternal antithyroid treatment causes goitre formation and bradycardia in the fetus, it is possible to give T ₄ by intra-amniotic injection. There may be biochemical evidence of transient hypothyroidism in clinically euthyroid babies of mothers treated with antithyroid drugs ( ). If fetal tachycardia suggests hyperthyroidism, treatment should be adjusted to maintain the fetal heart rate below 160 beats/min and careful assessment of fetal growth is necessary. Cordocentesis may be used to confirm the diagnosis.

Maternal hyperthyroidism – an approach to the assessment and management of children who are therefore at risk of neonatal hyper- or hypothyroidism. TFT, thyroid function tests; TSH, thyroid stimulating hormone; FT ₄ , free thyroxine.

Severe hyperthyroidism carries a high mortality in the newborn. The key to successful management is, first, anticipation and prevention, then control of thyroid status until the disease runs its self-limited course. If the fetus was hyperthyroid and the mother received a thionamide during pregnancy, then the wisest course will usually be to continue the thionamide medicine (CBZ is preferred in children: ) in a suitable neonatal dosage, which is carbimazole 0.5–1.5 mg/kg/24 h in divided doses every 8 hours. Regular assessment of thyroid function is necessary. Babies who are clinically euthyroid but subsequently show signs of hyperthyroidism will also need to be treated. Neonatal thyrotoxicosis usually remits after 2–5 months and so antithyroid medication can be withdrawn around this time.

Administration of a generous dose of antithyroid drug and simultaneous replacement with T ₄ (‘block and replace’) has been used in thyrotoxic infants but the relatively short duration of the hyperthyroid phase makes this approach impractical in most babies. If a baby is thought to be at particular risk of hyperthyroidism, if a sibling was symptomatic or if there is a high titre of maternal antibodies, then close clinical and biochemical surveillance should be continued for the first weeks of life. Treatment with an antithyroid preparation should be started promptly if necessary.

If acute thyrotoxicosis does occur, then in addition to a thionamide, propranolol 2.0 mg/kg/24 h in divided doses by mouth 6–8-hourly can be used to control the peripheral stimulatory effects of thyroid hormones and/or potassium iodide, as aqueous iodine solution (5% potassium iodide, 130 mg/ml) 0.05–0.1 ml 8-hourly can be administered to prevent synthesis and release of thyroid hormones from the gland. Radiographic iodine-containing agents such as sodium ipodate (0.5 mg every 3 days) have also been used in the treatment of neonatal Graves disease ( ). Severely ill babies can be treated with sedatives as well as a glucocorticoid such as prednisolone (2 mg/kg/day) which suppresses T ₄ to T ₃ conversion.

Occasionally an infant born to a mother with Graves disease will be found to have a hypothyroid picture with low thyroid hormone levels and a raised or suppressed TSH. In some instances, a low TSH and low T ₄ reflect suppression of the hypothalamopituitary–thyroid axis by the transplacental passage of thyroid hormone from a hyperthyroid mother, and these babies will need T ₄ replacement whilst the axis recovers ( ; ; ). The transplacental passage of blocking antibodies and exposure to antithyroid drug may account for low thyroid hormone levels.

The calcium-sensing receptor

The extracellular CaSR is one of a group of G protein-coupled cell surface receptors. The CaSR modulates the production and secretion of PTH by the chief cells in the parathyroid gland ( ). The CaSR has been termed the ‘calciostat’ ( ) because of its fundamental role in maintaining the narrow normal range ( ) of ionised calcium in blood. Rising calcium levels inhibit chief cell activity whilst falling levels of serum calcium will result in a rise in CaSR-induced cellular function. The CaSR is also known to be expressed in other tissues involved in calcium homeostasis, such as bone, gut and kidneys.

Gene defects that result in impaired or enhanced receptor activity can alter the ‘set-point’ and lead to hypo- or hypercalcaemia. For instance, AD familial benign hypocalciuric hypercalcaemia (impaired receptor function) and hypercalciuric hypocalcaemia (enhanced receptor function) may present in the neonatal period. These rare but important causes of neonatal hyper- and hypocalcaemia are described in more detail below.

Parathyroid hormone

PTH, a polypeptide with 84 amino acid residues, is secreted by the four (occasionally up to seven) parathyroid glands that usually lie behind the thyroid gland. The 34 amino-terminal residue has full biological activity and the function of the remainder of the polypeptide is unknown . Reduced calcium binding on the extracellular side of the CaSR results in an increase in PTH release via exocytosis from the chief cell of the parathyroid gland ( ). An increase in circulating phosphate will have a similar effect as it will complex with free ionised calcium to form calcium phosphate which is not detected by the CaSR. Low magnesium levels inhibit PTH secretion and so serum levels of this mineral should be measured whenever hypocalcaemia is being investigated.

When PTH is released, it is rapidly degraded into a number of fragments, the more active of which include the amino-terminal sequence. The half-life of PTH is a few minutes but some inactive fragments remain in the circulation longer.

PTH has three key actions:

• 1
  

  osteoclast stimulation with an associated release of calcium from bone

  
  
• 2
  

  to promote renal phosphate excretion (and sodium, potassium and bicarbonate) and decrease renal tubular calcium reabsorption (and magnesium and hydrogen ions)

  
  
• 3
  

  the stimulation of 1-α hydroxylase enzyme activity in the proximal renal tubule, thereby increasing the formation of 1,25-dihydroxyvitamin D. This will, in turn, promote intestinal calcium absorption.

The first two actions will regulate calcium levels acutely whilst the third has a more delayed onset of action.

Vitamin D

The two naturally occurring forms of vitamin D are ergocalciferol (D ₂ ), which is derived from ingested vegetables, and cholecalciferol (D ₃ ), which is formed from 7-dehydrocholesterol by the effects of ultraviolet light on the skin. Vitamin D ₃ is also absorbed from animal sources in the diet by the upper part of the small intestine. The formed cholecalciferol and absorbed ergo-/cholecalciferol are hydroxylated, primarily by the liver, to 25-hydroxyvitamin D, which is the major circulating vitamin D metabolite. This undergoes 1-hydroxylation in the kidney to 1,25-dihydroxyvitamin D, the metabolically active form, or to the less active 1,24,25-trihydroxyvitamin D or 24,25-dihydroxyvitamin D. 1-hydroxylation is stimulated and 24-hydroxylation inhibited by hypocalcaemia, hypophosphataemia and PTH. Only 1–3% of circulating vitamin D is free, with most bound to vitamin D-binding protein, an α ₂ -globulin.

The major effect of vitamin D is to increase serum levels of calcium and phosphate by facilitating their absorption from the gut. Vitamin D also enhances the mobilisation of calcium and phosphate from bone when dietary calcium is inadequate and promotes bone mineralisation by maintaining adequate concentrations of calcium and phosphate in the vicinity of unmineralised osteoid. Fetal 1,25-dihydroxyvitamin D is probably the major stimulus for the placental transfer of calcium.

Other peptide hormones with a more peripheral role in calcium homeostasis

Causes

The causes of hypocalcaemia are shown in Table 34.13 .

Treatment

Severe symptoms of hypocalcaemia, such as seizures, may be treated with 10% calcium gluconate 0.5–2 ml/kg (0.11–0.46 mmol/kg) by slow i.v. injection over 5–10 minutes. If necessary, this may be followed by continuous infusion of diluted 10% calcium gluconate, 2.5 ml/kg/24 h, with careful monitoring of cardiac rate and rhythm and serum calcium concentration. Calcium solutions are irritant to veins and should ideally be administered via a central venous catheter. Oral calcium supplements (0.25 mmmol/kg qds adjusted according to response) and vitamin D or a vitamin D analogue may be needed for long-term management.

Early hypocalcaemia

Symptomatic hypocalcaemia during the first 3 days of life occurs more commonly in preterm babies, infants of mothers with pre-eclampsia or diabetes and those who suffer birth asphyxia, sepsis or other perinatal stress. To some extent this can be viewed as an exaggeration of the normal postnatal fall in calcium levels. In the preterm infant this reflects a somewhat delayed PTH response to low calcium levels which is compounded by a relative renal resistance to the phosphaturic actions of of PTH. Many infants with early neonatal hypocalcaemia, especially those with diabetic mothers, also have low serum magnesium levels, which, like the hypocalcaemia, usually improve spontaneously.

Late hypocalcaemia due to inappropriate feeds

This clinical picture is now much less common than it was when unmodified cow’s milk formulas were widely used. It usually presented with seizures in apparently normal term infants on the fifth to 10th days of life. Classically this was due to the high phosphate and relatively low calcium content of cow’s milk. Increased absorption of phosphate caused hyperphosphataemia, which in turn depressed the serum calcium.

Hypoparathyroidism ( )

Hypoparathyroidism is a rare cause of hypocalcaemia in the newborn period. It usually presents after 5 days of age with overt signs of hypocalcaemia. Other biochemical findings include hyperphosphataemia, hypomagnesaemia and a normal or low alkaline phosphatase. The diagnosis is confirmed by the finding of low or absent immunoreactive PTH levels at the time of hypocalcaemia.

The condition may be familial, with X-linked recessive, AD and AR inheritance patterns described, and the underlying cause typically reflects an abnormality of gland development, abnormalities of the PTH molecule itself or abnormalities in other aspects of cellular function. Hypoparathyroidism may be an isolated disorder or it may occur in association with other abnormalities.

Other causes of hypocalcaemia

Treatment

Whilst the primary cause must be corrected if possible, the key component of short-term treatment of neonatal hypercalcaemia is a reduction in dietary calcium intake using low-calcium feed. Generous hydration and furosemide diuresis promote urinary calcium loss if a more acute reduction is needed. Glucocorticoids (e.g. hydrocortisone 1 mg/kg 6-hourly) reduce intestinal calcium absorption but the effect is slow and they should not be used if neoplasia is suspected. Calcitonin (10 U/kg i.v.) may be useful and has its maximum effect in 1 hour; infusions may be repeated 4-hourly. The hormone is antigenic and so the synthetic derivative salcatonin should be used in the longer term. Experience with the bisphosphonates is increasing ( ) and pamidronate in a dose of 0.5–1.0 mg/kg as an i.v. infusion may be worth considering in particularly difficult cases ( ). Long-term safety data following administration in infancy are not yet available.

Causes of hypercalcaemia ( Table 34.14 )