Congenital Hypothyroidism with a Delayed Thyroid-Stimulating Hormone Elevation in Very Premature Infants: Incidence and Growth and Developmental Outcomes

Congenital Hypothyroidism with a Delayed Thyroid-Stimulating Hormone Elevation in Very Premature Infants: Incidence and Growth and Developmental Outcomes Hyung Chul Woo, MD, Ariel Lizarda, BS, Richard Tucker, BA, Marvin L. Mitchell, MD, Betty Vohr, MD, William Oh, MD, and Chanika Phornphutkul, MD Objective To test the hypothesis that very low birth weight (VLBW) and extremely low birth weight (ELBW) infants have an increased incidence of congenital hypothyroidism (CH) with a delayed thyroid-stimulating hormone (TSH) elevation and that the outcomes of these infants are similar to control infants. Study design Retrospective analysis of newborn thyroid screening data for 92 800 live births in Rhode Island to identify CH with a delayed TSH elevation. Developmental, growth, and endocrine outcomes of the index cases were assessed at 18 months corrected age. Results CH with a delayed TSH elevation occurred in 1 in 58 ELBW, 1 in 95 VLBW, and 1 in 30 329 infants weighing $1500 grams (P < .0001). The incidence of head circumference <10th percentile was higher in VLBW infants with CH associated with a delayed TSH elevation (P < .05), and the mean head circumferences, weights, lengths, and developmental scores were similar to matched control infants. Three infants received short-term levothyroxine replacement. Conclusions The incidence of CH with a delayed TSH elevation was higher in ELBW and VLBW infants compared with infants weighing $1500 grams. The outcomes of these infants were comparable with matched control infants. (J Pediatr 2011;158:538-42). See editorial, p 525 and related article, p 532

C

ongenital hypothyroidism (CH) is detected by newborn screening (NBS). The estimated incidence has progressively increased in the past two decades.1-5 The most common form of CH is primary hypothyroidism, which is defined by low thyroxine (T4) and elevated thyroid-stimulating hormone (TSH) levels.6 Infants with CH are rarely symptomatic at birth. Delayed or inadequate treatment results in poor cognitive outcomes. The implementation of newborn screening for hypothyroidism has led to the early identification of infants with CH, prompt institution of treatment, and the consequent prevention of irreversible central nervous system damage.7-10 Given its public health relevance, the contributing factors to the increased incidence of CH were recently examined in detail.5 One of the contributing factors to the increased incidence of CH probably is preterm birth. In term infants, the physiological TSH surge occurs within the first 24 hours and subsequently declines.11,12 TSH remains elevated in patients with primary hypothyroidism, permitting the identification of CH when blood for the NBS is collected on day 2 to 3 of life. In contrast, some premature infants had an attenuated and delayed TSH elevation.6 They were also referred to as ‘‘atypical hypothyroidism.’’ The estimated incidence is 1 in 400 in infants weighing<1500 grams, which is significantly higher than term infants, and only one-third were identified by initial NBS.13 This observation led to the recommendation to perform repeat NBS in infants weighing <1500 grams in certain NBS programs.4,6,13,14 However, the practice is not universal, and long-term outcomes data are limited and presumably benign but not without controversies.15-18 Since earlier reports, there has been a progressive increase in survival of very low birth weight (VLBW) (birth weights between 1000 and 1499 grams) and extremely low birth weight (ELBW) (birth weights <1000 grams); hence, more NBS are done. We hypothesize that the incidence of delayed TSH elevation is high in the VLBW infants and even higher in ELBW infants. The primary objective of our study was to determine the incidence of CH with a delayed TSH elevation in VLBW and ELBW infants. A secondary aim was to examine the growth, neurodevelopmental, and endocrine outcomes

CH ELBW NBS TSH T4 VLBW

Congenital hypothyroidism Extremely low birth weight Newborn screening Thyroid-stimulating hormone Thyroxine Very low birth weight

From the Department of Pediatrics (H.C., R.T., B.V., W.O., C.P.), The Warren Alpert Medical School of Brown University, Providence, RI; Rhode Island Department of Health (A.L.), Providence, RI; and the University of Massachusetts (M.M.), Boston, MA Supported by the New England Regional Genetics Group (NERGG), University of New Hampshire (grant PZ09030). The authors declare no conflicts of interest. 0022-3476/$ - see front matter. Copyright ª 2011 Mosby Inc. All rights reserved. 10.1016/j.jpeds.2010.10.018

538

Vol. 158, No. 4  April 2011 of these infants at 18-months corrected age, an aspect that has not been documented systematically.

of Infant Development 2nd Edition Mental Developmental Index and Psychomotor Developmental Index.

Methods

Statistical Analysis Baseline characteristics of study groups were analyzed with t tests for continuous variables and c2 tests for categorical variables. Comparison of 18-month outcomes between groups was performed adjusting for the matched design. Analysis of continuous outcomes was done using hierarchical linear modeling to adjust for within-match group similarities. Analysis of categorical outcome was done with the Mantel-Haenszel test, using the matched groups as strata.

Institutional review board approvals were from the Women and Infants Hospital of Rhode Island, the Rhode Island Hospital – Lifespan Corporation, and the Rhode Island Department of Health. Using the Rhode Island electronic database (KIDSNET), which collected birth weights as well as NBS information of all newborn infants born in Rhode Island during the calendar years 2000 to 2006, a retrospective review was performed to identify VLBW (birth weight, 1000 to 1499 grams) and ELBW (birth weight <1000 grams) infants with CH caused by delayed TSH elevation. The Rhode Island NBS protocol measures both T4 and TSH simultaneously and defines an abnormal NBS for CH by a total T4 of #64.35 nmol/L or #5 mg/dL and an elevated TSH level. CH with a delayed TSH elevation was defined by a low or normal T4 and a normal TSH on the initial NBS, followed by an abnormally elevated TSH at a later period. NBS TSH cutoffs are age dependent: values $25 mIU/L are considered abnormal for infants <24 hours of age, and values of $20 mIU/L and $15 mIU/L are considered abnormal for infants 24 to 96 hours and >96 hours of age, respectively.13 Initial NBS for the infants were collected at
48 hours of age. In addition, for infants whose birth weights were <1500 grams, repeat NBS were also performed at 2, 6, and 10 weeks of age or until they reached 1500 grams. Repeat NBS were not done in infants $1500 grams unless medically indicated. For each VLBW or ELBW infant with CH and a delayed TSH elevation, the outcomes data from the Women and Infants Hospital Neonatal Follow-Up Clinic were matched to two control infants without CH or a delayed TSH elevation. Infants were matched on the following preset variables: gestational age 1 week, sex, birth weight 100 grams, antenatal steroids, postnatal steroids, intraventricular hemorrhage, mechanical ventilation $2 weeks, bronchopulmonary dysplasia, necrotizing enterocolitis, and an episode of bacterial sepsis. Information from the Women and Infants Follow-Up Clinic included the time of initiation and dosage of thyroxine used for treatment, thyroid function, growth variables (head circumference, weight, length) at birth, and at 18-months corrected age, and neurodevelopmental outcomes at 18 months controlled age, which consisted of the Bayley Scales

Results Using the Rhode Island KIDSNET database, 92 800 live births in Rhode Island were identified over the study period of 2000 to 2006. Of these, 1814 infants weighed <1500 grams, 885 had birth weights 1000 to 1499 grams, and 929 had birth weights <1000 grams. Three of the 885 VLBW infants (1 in 295 or 0.34%) and 16 of the 929 ELBW infants (1 in 58 or 1.72%) were identified to have CH with a delayed TSH elevation, compared with only 3 of the 90 986 infants weighing$1500 grams (1 in 30 328 or 0.0033%) (P < .0001, Table I). Newborn Screening Results and Patient Characterization The mean initial T4 was noted to be lower in VLBW and ELBW infants with delayed TSH elevation when compared with infants weighing $1500 grams (61 versus 104 nmol/ L). The initial T4 values were #64.4 nmol/L in 52.6% of VLBW and ELBW infants, and only 33.3% of infants weighing $1500 grams had low initial T4. Moreover, the peak TSH was >100 mIU/mL in 3 of the ELBW infants and 1 of the VLBW. None of the infants weighing $1500 grams had increases >40 mIU/mL. The mean age of infants weighing <1500 grams diagnosed with CH and a delayed TSH elevation was 21.9 days (Table II). Nineteen VLBW and ELBW infants were identified with CH with a delayed TSH elevation (Table III). The mean birth weight of the infants with a delayed TSH elevation was 790 grams and the mean gestational age was 25.9 weeks. History of dopamine/dobutamine exposure was documented in six of 12 patients who had delayed TSH elevation. History of topical iodine and urinary iodine content were not available. Treatment was initiated on three infants based on

Table I. Incidence of CH hypothyroid with delayed TSH elevation in Rhode Island newborn infants, 2000 to 2006 Birth weight (grams) Variable

<1000

1000-1499

<1500

‡1500

Live birth CH with delayed TSH surge Incidence in weight category P

929 16 1:58 (1.72%)

885 3 1:295 (0.34%)

1814 19 1:95 (1.05%)

90 986 3 1:30 329 (0.0033%)

<.0001 539

THE JOURNAL OF PEDIATRICS



www.jpeds.com

Table II. Characteristics of thyroid hormone abnormalities between VLBW/ELBW and infants with birth weight $1500 grams in infants with CH and a delayed TSH elevation Hypothyroid cases with delayed TSH elevation Mean age in days at initial TSH elevation Mean initial T4 (nmol/L)† Initial T4 <64.35 nmol/L, n (%) Mean peak TSH (mIU/L) Peak TSH >40 mIU/L, n (%) Peak TSH >100 mIU/L, n (%) Congenital cardiac disease, n (%)

VLBW and ELBW

‡1500 grams

19

3

22

25

61 10 (52.6) 62.28* 4 (21.1) 3 (15.8) 0 (0)

104 1 (33.3) 26.05 0 (0) 0 (0) 1 (33)

*One value was noted to be >500. †1 mg/dL = 12.87 nmol/L.

recommendations from pediatric endocrinologists. All had TSH >50 mIU/mL. The timing of discontinuation of treatment for these three patients as well as follow-up for transient TSH elevation was determined by the patients’ pediatric endocrinologists. The mean TSH of the 16 patients who were not treated was mildly elevated at 21 mIU/mL. The mean and median age of spontaneous resolution in untreated patients was 55 and 42 days, respectively. Growth and Neurodevelopmental Outcomes at Follow-up Three infants (infants 1, 6, and 13) died before discharge. Another 6 (infants 5, 7, 11, 17, 18, and 19) were lost to follow-up (Table III). Moreover, another infant had insufficient data at 18 months to be analyzed (infant 10). The ages of TSH resolution of infants who were lost to follow-up were similar to infants of whom we have long-term follow-up

Vol. 158, No. 4 information. Only 9 of 16 survivors (55%) available for neurodevelopmental follow-up were seen in the follow-up clinic at 18-months corrected age (infants 2, 3, 4, 8, 9, 12, 14, 15, and 16). Comparison of infants with CH with delayed TSH elevation with the control group showed no difference in gestational age, birth weight, birth length, birth head circumference, sex, or comorbidities, which included mechanical ventilation, intraventricular hemorrhage, antenatal steriods, bronchopulmonary dysplasia, necrotizing enterocolitis , and steroid exposure (data not shown). We were unable to control for dopamine/dobutamine exposure. Because maternal education and race have also been shown to affect neurodevelopmental outcomes, these factors were also evaluated independently between the two groups, and no statistical differences were identified (data not shown). Outcomes at 18-months controlled age revealed that 3 of 9 infants with CH and a delayed TSH elevation had head circumferences that were <10th percentile and none among the control infants (P < .014). Weight, length, and mental developmental index and psychomotor developmental index scores were similar to their control infants (Table IV).

Discussion Recent advances in the treatment of premature infants have led to a progressive reduction in their mortality and increased survival.18 The major risk factor associated with delayed TSH elevations include prematurity.13 Therefore, we sought to determine the incidence and outcome CH with a delayed TSH elevation in VLBW and ELBW infants compared with infants weighing $1500 grams. Our results concurred with previous reports that infants weighing <1500 grams have a higher

Table III. Characteristics of the 19 identified VLBW and ELBW infants with CH and a delayed TSH elevation At diagnosis of delayed TSH elevation At resolution of delayed TSH elevation Birth weight Gestational History of dopamine Patient (grams) age (weeks) Age (days) T4 (mg/dL) Serum TSH (mIU/mL) Age (days) T4 (mg/dL) Serum TSH (mIU/mL) exposure 1† 2 3 4 5* 6† 7* 8 9 10z 11* 12 13† 14 15 16 17* 18* 19* Average

490 490 500 570 580 610 650 750 750 800 820 830 860 890 930 965 1060 1174 1300 790.47

23 25 24 24 24 25 23 24 25 26 26 26 26 27 26 30 31 30 28 25.95

58 17 44 19 46 17 24 14 15 20 22 14 8 19 15 17 15 17 15 21.89

4.1 3.4 1.58 3.3 2.1 4.7 0.98 4.1 9.71 3.9 6.2 5.0 4.8 7.0 0.6 9.4 5.44 4.2 2.7

20 17.7 >500x 15.4 24.0 20.8 150x 15.3 28.4 22.3 15.7 17.1 16.1 18.3 197x 17.3 49.1 22.4 16.4

76 42 90 41 73 43 No value 38 24 45 34 42 25 60 42 No value 22 121 50 51.06

8.2 4.8 14.0 3.7 8.2 3.4 No value 4.9 6.37 3.5 7.0 10.4 3.7 9.4 8.4 No value 5.0 10.9 9.1

6.2 13.3 2.0 1.7 11.6 5.5 No value 0.8 5.14 9.8 5.1 10.1 0.9 7.4 1.0 No value 6.05 5.2 5.9

No history available No Yes Yes Yes Yes Yes No No history available No No Dobutamine exposure Yes No history available Yes No No No history available No history available

*Lost to follow-up. †Deceased. zInsufficient follow-up data at 18 months. xTreated.

540

Woo et al

ORIGINAL ARTICLES

April 2011

Table IV. Growth and neurodevelopmental outcomes of VLBW and ELBW infants with CH and delayed TSH surge at 18-months corrected age Variable Corrected gestational age* Length* <10th percentile Weight* <10th percentile Head circumference* <10th percentile Mental developmental index* <70 Psychomotor developmental index* <70 Neurological examination Normal Suspect Abnormal

CH with delayed TSH surge

Control

P

18.5  0.5 79.2  3.2 5 (28%) 10.3  1.4 8 (44%) 47.4  1.0 0 (0) 76  15.2 7 (39%) 77.6  11.5

.69 .66 .39 .29 .32 .14 .014 .83 .39 .73

3 (33%)

4 (22%)

.53

6 (67%) 2 (22%) 1 (11%)

14 (78%) 2 (11%) 2 (11%)

.71

18.4  0.6 78.6  2.9 4 (44%) 9.7  0.87 6 (67%) 46.8  1.6 3 (33%) 74.7  15.6 2 (22%) 75.9  15.0

P values are adjusted for matching. *Values are expressed as mean  SD.

incidence of CH with a delayed TSH elevation when compared with infants weighing $1500 grams. Given that repeat NBS was obtained in infants weighing $1500 grams when medically indicated based on risk factors, it is likely that our incidence of delayed TSH elevation in infants weighing $1500 grams was underestimated. However, our incidence of delayed TSH elevation in these infants is still higher than previously reported.13 The highest incidence of CH with a delayed TSH elevation occurred in ELBW infants (P < .0001) (Table I). The reason for this increased incidence may be due to a more attenuated hypothalamic-pituitary-thyroid response observed in ELBW infants.14 The mean T4 level was lower in the VLBW and ELBW infants when compared with infants weighing $1500 grams, which probably is explained by low thyroid-binding globulin, commonly observed in premature infants. The other explanation could be an increased incidence of sick euthyroid syndrome in very sick VLBW and ELBW infants. On the basis of the available data, we could not distinguish between the two conditions in the studied population. Given the continual increase in ELBW infant survival, we anticipate that the incidence of a delayed TSH elevation will continue to rise and remain a public health concern.5 The mean age of initial TSH elevation was
3 weeks. The timing of delayed TSH elevation was not significantly different from previous reports despite a much higher incidence in the ELBW infants in our study. This finding suggests that the risk of having a delayed TSH elevation occurs within the first month of an infants’ life when critically ill and possibly exposed to iodine during various procedures as well as dopamine/dobutamine. Dopamine exposure suppresses TSH secretion. Recent case series recommend obtaining repeat thyroid function in all infants exposed to dopamine.19 We then determined the natural history and course of the delayed TSH elevation, which has not been systematically documented. Of the 19 children with a delayed TSH eleva-

tion, thyroid hormone replacement was given by clinicians in 16% of the infants. Of these, the elevation of TSH values was minimal (mean 21 mIU/mL) and normalized spontaneously. In this study, despite the near normal T4 levels, the timing of TSH normalization varied greatly, with a median of 42 days. In children with severe primary CH, rapid normalization of TSH within the first 2 weeks of life correlates with better full-scale IQ,10 the long-term developmental impact of a mildly elevated TSH and normal T4 or subclinical/ compensated CH, treated or not, remains unclear. Infants with a delayed TSH elevation, who received levothyroxine therapy, interestingly, were not on therapy by the age of 18 months. This new observation suggests that thyroid dysfunction in VLBW and ELBW infants may well be mild or transient, which raises the question whether re-screening is necessary. A 1-year retrospective review of a Canadian experience from four major referral centers suggested that re-screening in VLBW and ELBW infants was unnecessary because all of the infants with CH were identified by the first specimens.15 The limitations of their observation, as noted by an accompanying editorial article,16 is the retrospective nature of the study and the small number of the population studied. Given that 16% of our infants had elevated TSH >50 mIU/mL, prompting initiation of therapy, it is likely that those infants benefited from standard of care treatment. Without repeating NBS in premature infants, these infants (3 of
1800) would have been missed based on the normal first screen and would not have received any therapeutic intervention. We hypothesized that the long-term outcomes of infants with delayed TSH elevation are similar to control infants. At 18-month follow-up, we observed similar weight and length in index cases versus the control infants. The Mental Developmental Index, Psychomotor Developmental Index, and neurological exams were also similar between the two groups. However, although the mean head circumference was not different from the control infants, there was an increased incidence of infants with a delayed TSH elevation who had head circumferences <10th percentile (P < .05). The significance of smaller head size is subject to intense debate. The predictive role of early head growth for subsequent outcomes was evaluated in 85 infants with birth weight <1500 grams.20 At 15 months of age, infants who were normocephalic at birth but had less postnatal head growth had intermediate Bayley scores.20 Our study subjects with delayed TSH elevation appeared to be born with normal head sizes, but the head circumferences showed a trend towards being smaller when compared with the control infants. This observation warrants longer follow-up in this group based on previous reports, despite reassuring Mental Developmental Index and Psychomotor Developmental Index at 18 months. n Submitted for publication May 31, 2010; last revision received Sep 18, 2010; accepted Oct 15, 2010. Reprint requests: Dr Chanika Phornphutkul, MD, The Warren Alpert Medical School of Brown University, Division of Pediatric Endocrinology and Metabolism, Department of Pediatrics, 593 Eddy Street, Providence, RI 02903. E-mail: [email protected]

Congenital Hypothyroidism with a Delayed Thyroid-Stimulating Hormone Elevation in Very Premature Infants: Incidence and Growth and Developmental Outcomes

541

THE JOURNAL OF PEDIATRICS



www.jpeds.com

References 1. Harris KB, Pass KA. Increase in congenital hypothyroidism in New York State and in the United States. Mol Genet Metab 2007;91:268-77. 2. United States General Accounting Office. Newborn Screening: Characteristics of State Programs. 2003. 3. Rose SR, Brown RS, Foley T, Kaplowitz PB, Kaye CI, Sundararajan S, et al. Update of newborn screening and therapy for congenital hypothyroidism. Pediatrics 2006;117:2290-303. 4. Maniatis AK, Taylor L, Letson GW, Bloch CA, Kappy MS, Zeitler P. Congenital hypothyroidism and the second newborn metabolic screening in Colorado, USA. J Pediatr Endocrinol Metab 2006;19:31-8. 5. Olney RS, Grosse SD, Vogt RF Jr. Prevalence of congenital hypothyroidism: current trends and future directions: workshop summary. Pediatrics 2010;125(Suppl 2):S31–S36. 6. 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. 7. Effects of neonatal screening for hypothyroidism: prevention of mental retardation by treatment before clinical manifestations: New England Congenital Hypothyroidism Collaborative. Lancet 1981;2:1095-8. 8. Hyman SJ, Greig F, Holzman I, Patel A, Wallach E, Rapaport R. Late rise of thyroid stimulating hormone in ill newborns. J Pediatr Endocrinol Metab 2007;20:501-10. 9. Elementary school performance of children with congenital hypothyroidism: New England Congenital Hypothyroidism Collaborative. J Pediatr 1990;116:27-32. 10. Selva KA, Harper A, Downs A, Blasco PA, Lafranchi SH. Neurodevelopmental outcomes in congenital hypothyroidism: comparison of initial T4 dose and time to reach target T4 and TSH. J Pediatr 2005;147:775-80.

542

Vol. 158, No. 4 11. DeLong GR, Robbins J, Condliffe PG. Iodine and the Brain. New York: Plenum Press; 1989. 12. Murphy N, Hume R, van Toor H, Matthews TG, Ogston SA, Wu SY, et al. The hypothalamic-pituitary-thyroid axis in preterm infants; changes in the first 24 hours of postnatal life. J Clin Endocrinol Metab 2004;89:2824-31. 13. Larson C, Hermos R, Delaney A, Daley D, Mitchell M. Risk factors associated with delayed thyrotropin elevations in congenital hypothyroidism. J Pediatr 2003;143:587-91. 14. Hunter MK, Mandel SH, Sesser DE, Miyabira RS, Rien L, Skeels MR, et al. Follow-up of newborns with low thyroxine and nonelevated thyroid-stimulating hormone-screening concentrations: results of the 20-year experience in the Northwest Regional Newborn Screening Program. J Pediatr 1998;132:70-4. 15. Vincent MA, Rodd C, Dussault JH, Van Vliet G. Very low birth weight newborns do not need repeat screening for congenital hypothyroidism. J Pediatr 2002;140:311-4. 16. Rapaport R. Thyroid function in the very low birth weight newborn: rescreen or reevaluate? J Pediatr 2002;140:287-9. 17. Reuss ML, Paneth N, Pinto-Martin JA, Lorenz JM, Susser M. The relation of transient hypothyroxinemia in preterm infants to neurologic development at two years of age. N Engl J Med 1996;334:821-7. 18. Fisher DA. Thyroid function and dysfunction in premature infants. Pediatr Endocrinol Rev 2007;4:317-28. 19. Filippi L, Pezzati M, Cecchi A, Poggi C. Dopamine infusion: a possible cause of undiagnosed congenital hypothyroidism in preterm infants. Pediatr Crit Care Med 2006;7:249-51. 20. Gross SJ, Oehler JM, Eckerman CO. Head growth and developmental outcome in very low-birth-weight infants. Pediatrics 1983; 71:70-5.

Woo et al