REGULAR ARTICLE

Permanent and transient congenital hypothyroidism in preterm infantsRamesh Srinivasan1, Sundeep Harigopal2, Steve Turner3, Tim Cheetham (tim.cheetham@nuth.nhs.uk)1,4

1.Department of Paediatric Endocrinology, The Great North Children’s Hospital, Newcastle Upon Tyne, UK2.Neonatal Intensive Care Unit, The Great North Children’s Hospital, Newcastle Upon Tyne, UK3.Department of Clinical Biochemistry, The Great North Children’s Hospital, Newcastle Upon Tyne, UK4.Institute of Genetic Medicine, Newcastle University, Newcastle Upon Tyne, UK

KeywordsCongenital hypothyroidism, Prematurity, Thyroid-stimulating hormone, Transient hypothyroidism

CorrespondenceDr Tim Cheetham, Department of Paediatric Endo-crinology, Newcastle University, The Great NorthChildren’s Hospital, Newcastle Upon Tyne, NE1 4LP,UK.Tel: +0044 191 282 9562 |Fax: +0044 191 2825485 |Email: tim.cheetham@nuth.nhs.uk

Received22 September 2011; revised 18 November 2011;accepted 18 November 2011.

DOI:10.1111/j.1651-2227.2011.02536.x

ABSTRACTAim: Transient fluctuations in thyroid function are well recognized in preterm infants.

We wanted to assess TSH variation in babies with transient and permanent congenital

hypothyroidism (CHT).Methods: Whole bloodspot TSH data in preterm infants (<35 weeks; 2005–2010)

were assessed, and infants with bloodspot TSH values >6 mU ⁄ L identified. Permanent

CHT was defined as a requirement for thyroxine beyond 3 years of age.Results: A first TSH sample was obtained from 5518 infants (median gestational

age, 32 w; range, 22–35), with a second sample obtained from 5134 infants (median

gestational age, 32 w; range, 22–35). Five infants had raised TSH concentrations on both

occasions. Three of the five infants had a serum TSH >80 mU ⁄ L on second screen but two

came off thyroxine beyond 3 years of age. All preterm babies with permanent or transient

hypothyroidism were detected by the first TSH cut-off of 6 mU ⁄ L. Only one infant with a

birth weight <1500 g remains on thyroxine treatment beyond 2 years of age.Conclusions: The incidence of permanent CHT in preterm infants is similar to term

infants. Profound abnormalities of thyroid function can occur in preterm babies with

transient hypothyroidism but both categories of hypothyroidism can be detected by a

‘once-only’ TSH screening strategy with a relatively low cut-off.

INTRODUCTIONNational newborn screening for congenital hypothyroid-ism (CHT) was introduced in the UK in 1981 (1) withrevised UK neonatal screening guidelines (2005) recom-mending that the preterm infant should have a secondblood spot sample for measurement of thyroid-stimulatinghormone (TSH) taken at the equivalent of 36 weeks gesta-tion (2). In practice, this means all babies born before35 weeks gestation are re-screened because older babieswill be at or beyond the equivalent of 36 weeks when thefirst sample is taken and so will not be eligible for the sec-ond test. This strategy reflected a concern that some pre-term infants might be unable to mount an appropriateTSH response in the presence of an abnormal thyroidgland (3) – a pattern that has been described as atypicalhypothyroidism (4,5). It is recognized that TSH fluctua-tions that result in an infant screening positive can reflecttemporary factors such as exposure to topical iodine (6,7)or drugs (8,9). These factors are more likely to be a con-sideration in the preterm infant. Hence, preterm babieswith an increased TSH may have transient rather thanpermanent hypothyroidism (10) and not require long-termthyroxine therapy. Both transient and permanent CHTcould fall under the ‘atypical hypothyroid’ umbrella. Theincreased prevalence of transient CHT may account forthe reported increase in CHT in the preterm (11,12).

Whether babies with transient hypothyroidism benefitfrom intervention is unknown (13).

Many of the reports of preterm screening programmesfocus on babies under 1500 g but all preterm infants are cur-rently re-screened in the UK and not just those with a weightbelow a specific threshold. We have previously reported theresults of a preterm screening programme Korada et al. (14)and in our current study we wanted to focus on the naturalhistory of the fluctuations in TSH that occur in early life in alarger group of babies. We also wanted to see whether theTSH threshold in our region of the UK would detect bothpermanent CHT and transient hypothyroidism.

METHODEthics approval was obtained for these studies. The North-east and North Cumbria region of England has on average

Key notes• Profound abnormalities of thyroid function can occur in

preterm babies who do not have permanent CHT. Alower TSH cut-off, using a contemporary assay, candetect both transient and permanent CHT without theneed for a second screen.

Acta Pædiatrica ISSN 0803–5253

ª2011 The Author(s)/Acta Pædiatrica ª2011 Foundation Acta Pædiatrica 2012 101, pp. e179–e182 e179

35 000 live births per year. Whole blood TSH in preterminfants (<35 week gestation) was obtained from the North-ern Regional Newborn screening database over a period of5 years (April 2005–March 2010).

Blood spot–screening protocolThe National recommended cut-off for investigation into(CHT) in the UK is a blood spot TSH of 10 mU ⁄ L (2) butour regional threshold (north-east England) is 6 mU ⁄ L.This reflects the fact that TSH values around 10 mU ⁄ L inearlier assays equated to TSH values of 6 mU ⁄ L with thenew method. Infants with a blood spot (TSH) concentra-tion >20 mU ⁄ L are considered to have a positive screeningresult. Infants with an initial blood spot TSH concentra-tion between 6 mU ⁄ L and 20 mU ⁄ L are considered tohave a ‘borderline’ result, and the assay is repeated on ablood spot from the same card. If the concentrationremains >6 mU ⁄ L on re-assay, a repeat blood spot sampleis collected. If the repeat blood spot result is ‡6 mU ⁄ L,then this is considered a positive result and the infantreferred for further investigation. Preterm infants thenundergo a second test at a corrected age of 36 weeks usingthe same 6 mU ⁄ L TSH cut-off.

Permanent versus transient CHTPermanent CHT can be defined by a requirement for thyr-oxine beyond 3 years of age. However, it would be possibleto refine this further by stating that unequivocal permanentCHT is defined by an ongoing need for T4 beyond 3 years ofage with a baseline serum TSH that is > 10 mU ⁄ L. TransientCHT can be defined as an elevated serum TSH > 10 mU ⁄ Lin early life with no long-term requirement for T4 therapybeyond 3 years of age.

Assay performanceInterassay coefficients of variation for blood spot TSHassays were 11% and 12% for the Dissociation-EnhancedLanthanide Fluorescent Immunoassay (DELFIA) (PerkinElmer, Boston, MA, USA) fluoroimmunometric assay atTSH values of 16 mU ⁄ L and 60 mU ⁄ L, respectively. Bloodspot TSH values are typically 40–50% of serum concentra-tions.

RESULTSA first TSH sample was obtained from 5518 infants (mediangestational age, 32 weeks; range, 23–35), with a secondsample obtained from 5134 (median gestational age,

32 weeks; range, 23–35). In 38 infants, the TSH concentra-tion fell from above to below the screening threshold and in6 infants values rose from below the screening threshold to6–10 mU ⁄ L on the second screen. However, TSH concen-trations fell below 6 mU ⁄ L in five of these infants when theywere subject to the second screen and the remaining babyhad a serum TSH of 6.8 mU ⁄ L with a normal free thyroxine.

Five infants had raised TSH concentrations (>6 mU ⁄ L)on both occasions (birth weight 830–2810 g). Three of thefive infants had a biochemical picture that was thoughthighly suggestive of true CHT (serum TSH >80 mU ⁄ L onsecond screen; Table 1).

Infant 1 had an initial TSH concentration of 8.5 mU ⁄ L(subsequent serum TSH, 134.7 mU ⁄ L; free thyroxine,10.4 pmol ⁄ L; birth weight, 2810 g). She was treated withthyroxine but is now off treatment after a trial off therapy atthe age of 3 years. She did not undergo thyroid imaging.

Infant 2 had an initial TSH concentration of 12.2 mU ⁄ L(subsequent serum TSH, 86.7 mU ⁄ L; free thyroxine,11 pmol ⁄ L; birth weight, 1710 g). She was treated with thy-roxine but is now off treatment after a trial off therapy at theage of 4 years. Her thyroid ultrasound examination wasnormal.

Infant 3 had an initial TSH concentration of 21.6 mU ⁄ L(subsequent serum, TSH 303 mU ⁄ L; free thyroxine,12.5 pmol ⁄ L; birth weight, 1910 g). She was treated withthyroxine and is still on therapy at the age of 4 years. Shehas not undergone thyroid gland imaging to date.

Infant 4 had an initial TSH concentration of 6.8 mU ⁄ L(subsequent serum TSH, 19.7 mU ⁄ L; free thyroxine,16 pmol ⁄ L; birth weight, 830g). He was treated with thyrox-ine and is still on therapy at the age of 2 years. His thyroidultrasound examination was normal.

Infant 5 had an initial TSH concentration of 11 mU ⁄ L(subsequent serum, TSH 8.3 mU ⁄ L; free thyroxine,13 pmol ⁄ L; birth weight, 2760 g). He was treated withthyroxine and is still on therapy at the age of 18 months.His thyroid ultrasound examination was normal.

On the basis of these data, the incidence of permanentCHT in the population of preterm infants is at most 3 per5100 infants. However, unequivocal CHT with a confirmedrequirement for long-term replacement has an incidence of1 per 5100 infants.

DISCUSSIONOur experience has highlighted a number of highly topicalissues including the problem of CHT definition.

Table 1 Preterm infants with first and second blood spot thyroid-stimulating hormone (TSH) >6 mU ⁄ L

Infant Sex Gestation TSH1 (mU ⁄ L) TSH2 (mU ⁄ L) Serum TSH (mU ⁄ L) Free T4 (pmol ⁄ L) BW and patient details

1 F 35 8.5 11.8 134.7 10.4 2.8 kg. Off thyroxine at 3 years of age

2 F 33 12.2 126 86.7 11 1.7 kg. Off thyroxine at 4 years of age

3 F 33 21.6 188 303 12.5 1.9 kg. On thyroxine 50 ⁄ 75 mcg alternate days at 4 years of age

4 M 32 6.8 7 19.7 16.3 0.835 kg. On thyroxine 30 mcg at 2 years of age

5 M 34 11 7.6 8.3 13 2.7 kg. On thyroxine 40 mcg at 18 months of age

TSH in preterm infants Srinivasan et al.

e180 ª2011 The Author(s)/Acta Pædiatrica ª2011 Foundation Acta Pædiatrica 2012 101, pp. e179–e182

A second screening test will detect a small number ofbabies who demonstrate a profound increase in TSH fol-lowing an initial blood spot that is in the borderline rangebetween 6 and 20 mU ⁄ L on a bloodspot TSH assay (10,14).Most of the repeat testing literature has focused on a birth-weight cut-off such as 1500 g (15–17) but our data confirmthat major fluctuations can occur in infants with a birthweight greater than this.

Whilst the initial conclusion might be that repeat testingis therefore justified, our data have confirmed that some ofthese babies do not have permanent CHT as reflected by thefact that they do not require long-term thyroxine treatment.A profound increase in TSH does not, therefore, necessarilyindicate permanent CHT – here defined by the need forlong-term thyroxine supplementation (18). Even serumTSH values >80 mU ⁄ L in early life do not necessarily reflecta requirement for long-term thyroxine therapy. Whetherbabies with such profound but transient alterations in TSHbenefit from temporary thyroxine therapy is unknown.

Although biochemistry is the most important parameterwhen deciding upon which babies to treat with thyroxine,information about underlying thyroid gland morphologycan be useful when advising families about the need forlonger term treatment. We did not have the comprehensiveimaging data required to establish the precise aetiology inall of these babies although the only infant requiring long-term replacement, who has not been imaged, has requiredan increasing dose of thyroxine replacement. The fact thatimaging is not an essential component of management andthe fact that it does not usually influence the decision to ini-tiate treatment has been highlighted in a recent review (19).Another potential limitation of our data is the fact thatbabies were managed in different centres by different clini-cians and hence we cannot be certain that the criteria forreassessment and for withdrawing thyroxine therapy werethe same.

Although our population of preterm infants is not partic-ularly large, the incidence of unequivocal permanent CHTappears to be similar to figures for CHT as a whole (20).

In summary, our data have shown that profound abnor-malities of thyroid function can occur in preterm babieswho do not have permanent CHT. These abnormalities canoccur in babies with a birth weight >1500 g. The extent ofthese fluctuations helps to explain why figures documentingthe incidence of CHT in preterm infants can be relativelyhigh in comparison with those for CHT in term infants (21).

Our data underline the importance of all babies undergo-ing a formal trial off T4 therapy at around 3 years of age(13,22) if TSH values have not been elevated whilst on ther-apy beyond the first few months of life. Finally, it is possiblethat a lower TSH cut-off, using a contemporary assay, willdetect both transient and permanent forms of CHT withoutthe need for a second screen.

CONFLICT OF INTEREST AND FUNDINGNone.

References

1. Department of Health and Social Security (DHSS). JointStanding Committee on Screening in Medical Care. Screeningfor early detection of congenital hypothyroidism. London:DHSS, 1981: HN (81) 20.

2. UK Newborn Screening Programme Centre. Newborn bloodspot screening in the UK. Policies and Standards. 2005. Avail-able at: http://newbornbloodspot.screening.nhs.uk/cht.

3. Fisher DA. Thyroid function and dysfunction in prematureinfants. Pediatr Endocrinol Rev 2007; 4: 317–28.

4. Mandel SJ, Hermos RJ, Larson CA, Prigozhin AB, Rojas DA,Mitchell ML. Atypical hypothyroidism and the very low birthweight infant. Thyroid 2000; 10: 693–5.

5. Frank JE, Faix JE, Hermos RJ, Mullaney DM, Rojan DA, Mitch-ell ML, et al. Thyroid function in very low birth weight infants:Effects on neonatal hypothyroidism screening. J Pediatr 1996;128: 548–54.

6. Linder N, Davidovitch N, Reichman B, Kuint J, Lubin D, Meye-rovitch J, et al. Topical iodine-containing antiseptics and sub-clinical hypothyroidism in preterm infants. J Pediatr 1997; 131:434–9.

7. Larson C, Hermos R, Delaney A, Daley D, Mitchell M. Risk fac-tors associated with delayed thyrotropin elevations in congeni-tal hypothyroidism. J Pediatr 2003; 143: 587–91.

8. Filippi L, Cecchi A, Rubaltelli FF, Tronchin M, Dani C, PezzatiM, Seminara S, et al. Dopamine infusion and hypothyroxina-emia in very low birth weight preterm infants. Eur J Pediatr2004; 163: 7–13.

9. Re RN, Kourides IA, Ridgway EC, Weintraub BD, Maloof F.The effect of glucocorticoid administration on human pituitarysecretion of thyrotropin and prolactin. J Clin Endocrinol Metab1976; 43: 338–46.

10. Woo HC, Lizarda A, Tucker R, Mitchell ML, Vohr B, Oh W,et al. Congenital hypothyroidism with a delayed thyroid-stimu-lating hormone elevation in very premature infants: incidenceand growth and developmental outcomes. J Pediatr 2011; 158:538–42.

11. Hinton CF, Harris KB, Borgfeld L, Drummond-Borg M, EatonR, Lorey F, et al. Trends in incidence rates of congenital hypo-thyroidism related to select demographic factors: data from theUnited States, California, Massachusetts, New York, and Texas.Pediatr 2010; 125: S37–47.

12. Parks JS, Lin M, Grosse SD, Hinton CF, Drummond-Borg M,Borgfeld L, et al. The impact of transient hypothyroidism onthe increasing rate of congenital hypothyroidism in the UnitedStates. Pediatr 2010; 125: S54–63.

13. Buyukgebiz A. Newborn screening for congenital hypothyroid-ism. J Pediatr Endocrinol Metab 2006; 19: 1291–8.

14. Korada M, Pearce MS, Ward Platt MP, Avis E, Turner S,et al. Repeat testing for congenital hypothyroidism in pre-term infants is unnecessary with an appropriate thyroid stim-ulating hormone threshold. Arch Dis Child Fetal NeonatalEd 2008; 93: F286–8.

15. Wiley V, Bijarnia S, Wikcken B. Screening for hypothyroidismin very low birth weight babies. Rev Invest Clin 2009; 61: 31–6.

16. Mengreli C, Kanaka-Gantenbein C, Dacou-Voutetakis C, et al.Screening for Congenital Hypothyroidism: the Significance ofThreshold Limit in False-Negative Results. JCEM 2010; 95:4283–90.

17. Bijarnia S, Wilcken B, Wiley VC. Newborn screening for con-genital hypothyroidism in very-low-birth-weight babies: theneed for a second test. J Inherit Metab Dis 2011; 34:827–33.

18. Ng SM, Anand D, Weindling AM. High versus low dose of ini-tial thyroid hormone replacement for congenital hypothyroid-ism. Cochrane Database Syst Rev 2009; 1: CD006972.

Srinivasan et al. TSH in preterm infants

ª2011 The Author(s)/Acta Pædiatrica ª2011 Foundation Acta Pædiatrica 2012 101, pp. e179–e182 e181

19. LaFranchi SH. Approach to the diagnosis and treatment ofneonatal hypothyroidism. J Clin Endocrinol Metab 2011; 96:2959–67.

20. Rastogi MV, LaFranchi SH. Congenital hypothyroidism. Or-phanet J Rare Dis 2010; 5: 17.

21. Harris KB, Pass KA. Increase in congenital hypothyroidism inNew York State and in the United States. Mol Genet Metab2007; 91: 268–77.

22. Yang RL, Zhu ZW, Zhou XL, et al. Treatment and follow-up ofchildren with transient congenital hypothyroidism. J ZhejiangUniv Sci B 2005; 6:1206–9.

TSH in preterm infants Srinivasan et al.

e182 ª2011 The Author(s)/Acta Pædiatrica ª2011 Foundation Acta Pædiatrica 2012 101, pp. e179–e182