Neonatal Screening for Congenital Hypothyroidism with Focus on Developing an Indian Screening Programme

Neonatal screening for congenital hypothyroidism, along with eradication of iodine deficiency in large parts of the world, has made it possible to prevent the development of permanent neurological impairment due to thyroid hormone deficiency in the developing brain. The first successful screening programme was demonstrated in Canada in 1973 and since then it has been standard of care in most developed societies. In India there is no national programme for neonatal screening, and screening is only done in selected larger hospitals on newborns whose parents fund it. This review summarises the current understanding of the various strategies for newborn screening that could potentially be employed in India with resource constraints. Once a case is detected, the further evaluation and determination of etiology is summarised. Treatment and long term follow-up with levothyroxine replacement is also described in detail as per current understanding.

The human foetus is capable of producing thyroid hormones at around 20 weeks of gestation. Even in the rare event that there happens to be a defect in thyroid organogenesis or an inborn error in thyroid hormone synthesis, the developing foetal brain is protected by the trans-placental supply of maternal T 4 . Within the cerebral cortex there is up-regulation of type 2 de-iodinase activity (Figure 1), leading to an enhanced supply of active T 3 to the cortex protecting the foetal brain from any significant neurological impairment. 1 This understanding underlies the remarkable success of neonatal screening programmes for thyroid dysfunction at birth in providing a good neurological prognosis for infants born with congenital hypothyroidism (CH).

Epidemiology of congenital hypothyroidism
The incidence of CH in unscreened populations, when the diagnosis is made clinically, is around one in 6000 births. 2 Newborn screening was first initiated in the Canadian province of Quebec in 1972 and in over 3 years, seven cases were detected among the 47,000 newborns screened. 3 Subsequently, incidence figures from around the globe have suggested one case of CH among 3,000-4,000 newborns screened. 4 However in the last 3 decades there has been an increase in case detection rates with current estimates of incidence rates between 1:1,400 to 1:2,800 newborns screened. 5 The primary reason for this increase in incidence of CH may relate to the change in screening strategies (from Free T 4 [FT 4 ] to thyroid stimulating hormone [TSH]) and to the lowering of thresholds for diagnosis. Most of the additional cases identified have milder forms of CH and the incidence rates for severe CH remain the same as in the decades preceding. Another factor in the apparent increase in CH rates is probably the change in demography in most Western populations with higher percentages of Asian patients in Western screening data. Asian newborns have higher rates of thyroid dyshormogensis. 6 A last factor contributing to the increase in the incidence rates could be related to the improved survival of preterm infants over the last three decades. 6 Indian data about the incidence of CH is scarce. Screening for CH among 40,000 newborns at the Wadia Hospital for Children in Mumbai revealed an incidence rate of 1 in 2,640 births. 7 Details of incidence rates published from other parts of the country are given in Table 1. [7][8][9][10]

Aetiology of congenital hypothyroidism
A vast majority of babies born with CH have an abnormal development of the thyroid gland.
Thyroid dysgenesis (abnormal thyroid gland development) accounts for 85% of all the cases of CH and is largely isolated and non-syndromic. Thyroid ectopy, hypoplasia and athyreosis are the three types of thyroid dysgenesis. Among the three, thyroid ectopy accounts for over 60% of cases and is much more common in girls than boys. Syndromic thyroid dysgenesis accompanies monogenic disorders of thyroid transcription factors (TTF)-1, TTF-2 and PAX-8 that lead to abnormal or absent thyroid development. 11 Details of the various permanent and transient aetiologies of CH in the newborns are given in Table 2

Screening strategies for congenital hypothyroidism
Screening for CH is complicated by the dramatic changes in TSH and thyroid hormone levels at birth and in the first month after birth. The magnitude of these hormonal changes differ in a preterm infant, small for gestational age (SGA) baby and a normal term appropriate for gestational age (AGA) baby, making single measurements difficult to interpret without appropriate local cut offs. The major changes in thyroid hormones at birth include: • A 50% increase in total T 4 which peaks on day 7 of infancy and stabilises by day 28 • A gradual three-to four-fold increase in total T 3 levels which reach adult levels by the end of 28 days of life • An abrupt increase in TSH within 6 hours of delivery compared to the cord blood TSH values followed by a dramatic fall in the first seven days of life. 12 Because of these changes the optimal time for screening infants for CH is probably a week after birth. However, by this time, babies would have usually been discharged from hospital and collection of samples from the community would entail extra expenditure and effort. This is the reason why most Western protocols suggest that screening for healthy newborns should be done between day 2 and 5 after birth prior to discharge from the hospital. For critically ill, preterm or home delivered infants, samples can be collected at the end of the first week. In India, because of the Relative normal brain T3 concentrations  Primary TSH testing between days 2-5 using heel prick This is by far the commonest protocol used in Western countries and the current screening test of choice in the recent guideline by major worldwide paediatric endocrine organisations. 13 The exceptions are some areas in the United States and Netherlands. TSH is tested in a central laboratory from the blood spots dried on filter paper. The sample protocol followed by the National Health Service in Scotland is described in Figure 2. The disadvantages with this approach are: • Rare patients with central hypothyroidism (Incidence 1:45000 babies) will be missed • Isolated hypothyroxinemia • Transient hypothyroxinemia of prematurity will not be detected and • Patients with thyroglobulin deficiency and delayed rise in TSH will be missed Problems with delayed TSH rise are also seen in preterm babies and very low birth weight babies. This approach usually has a recall rate of two babies for every baby with CH who is identified. 14 Moreover when resource constraints are encountered as in India, primary TSH testing will help identify most cases of mild to severe primary hypothyroidism.
A second screening should be considered in preterm, low birth weight and acutely unwell neonates. 13

TSH testing in cord blood
TSH testing in cord blood is likely to encounter lots of false alarms.
However, this is the easiest sample to be obtained at the time of active delivery. A protocol followed at Christian Medical College, Vellore is described in Figure 3.

Primary T 4 (total) testing followed by TSH if required
An alternate to the above is primary total T 4 testing followed by TSH estimation in the same sample when the measured total T 4 is below a defined threshold (e.g., below -0.8 standard deviation [SD] of the days mean T 4 ). 15 To improve the sensitivity of total T 4 testing, clinicians in the Netherlands also measure thyroid binding globulin in the filter paper sample when T 4 is -1.6 SD or less. Using this method, the Netherlands group showed an incidence rate of central hypothyroidism in 1 out of 20,263 babies screened. 16 Despite the favourable reports of this approach in Europe the first pathway of primary TSH testing is the one which is widely followed. The major disadvantages of this method are subclinical hypothyroidism will be missed, and delayed TSH rise with normal T 4 will be missed. However, the main disadvantage of this protocol is the much larger number of patients that need to be recalled (around 12 babies) to diagnose one baby with CH. 17

Further evaluation of babies diagnosed with congenital hypothyroidism Clinical Assessment
Babies referred with abnormal screening tests should be seen promptly in specialist clinics. Clinical history would focus on symptoms of  Additionally, cardiac examination and hip examination should be done as there is an increased prevalence of these disorders with CH. 13,18 Dysmorphic features should be looked for to consider syndromes associated with transient and permanent CH. These features are described in Table 2.

Investigations
At least 1 ml of venous blood should be collected for thyroid function testing and an additional 1 ml should be drawn to estimate thyroglobulin. If there is a maternal family history of thyroid disease then samples should be drawn from the mother and baby for antithyroid peroxidase (TPO) antibodies and TSH-R blocking antibodies.
In cases where there is parental consanguinity, DNA samples can be collected from the parents and the baby.
Some centres would routinely do a knee x-ray in babies with CH to document the bone retardation. This correlates with the degree of intra-uterine hypothyroidism and may predict future intelligent quotient (IQ) outcome and is recommended in the current guideline. 13,19 Diagnostic imaging

Further investigations
Ectopic thyroid and athyreosis with undetectable serum thyroglobulin leads to straight forward confirmation of diagnosis. Normal or hypoplastic glands with reduced isotope uptake should prompt investigations for maternal antibodies (anti-TPO and TSH-R). DNA analysis for TSH-R mutations can be undertaken if available and if there are no maternal antibodies to account for the low uptake.
A large or normal sized gland with normal or increase isotope uptake is suggestive of dyshormonogenetic CH. If iodine insufficiency is suspected then urine iodine excretion should be determined in both mother and child. The remaining causes of dyshormonogenesis are inherited enzyme defects, and require mutational analysis.

Treatment and follow up Treatment
Replacement with levothyroxine should be administered as soon as the diagnosis is established and should not be needlessly postponed for investigations to determine the aetiology of CH.
The aim of therapy is to render the newborn euthyroid as early as possible to prevent any permanent neurological sequelae. There is a direct inverse relationship between the age of starting treatment and the IQ of the child. 21  The goal of therapy is to normalise FT 4 and TSH within a fortnight of commencing therapy. Longer time periods for correction are associated with poorer neurological outcomes. 23 The starting doses advised are 50 µgm/day of levothyroxine for babies who weigh ≥2.5 kg and a dose of 15 µg/kg/day for babies who weigh less than 2.5 kg. In first months, thyroid functions are done weekly. 20  Auxological and neurodevelopmental outcomes in treated children Current guidelines suggest that parents and patients should be reassured about normal growth, puberty and fertility with appropriate and adequate treatment. 13 Children treated in the early decades of CH screening exhibited subtle neurodevelopmental and physical impairments and were also reported to have a reduced quality of life. 25 With current practice changes and early treatment these impairments are unlikely to happen.
Most children with CH regardless of severity are likely to have neurocognition and behaviour on an average within the normal limits. 17 However a few children with CH are likely to face clinically significant learning disability. 26,27 Conclusion The overall outcomes of babies born with CH are good. Early and high dose levothyroxine with normalization of TSH within 2 weeks of therapy has improved neurological outcomes from previously. Screening protocols need to be planned as per local circumstances, but a clear plan needs to made for further evaluation and commencing early treatment with levothyroxine. In India, the logistics for universal screening are likely to be enormous and the cost required for this is not likely to be available in the near future. Most screening is done in targeted populations at individual hospitals were the parents pay for the screening. A uniform protocol to be followed across the country is also likely to be difficult considering the different patterns of health care access across the country. q