PEDIATRICS Vol. 115 No. 2 February 2005, pp. 406-410 (doi:10.1542/peds.2004-0192)
Hypothyroxinemia in Mechanically Ventilated Term Infants Is Associated With Increased Use of Rescue Therapies
,¶,||
* Department of Pediatrics Thomas Jefferson University, Philadelphia, Pennsylvania
Department of Pediatrics, duPont Hospital for Children, Wilmington, Delaware
Department of Neonatology, North Shore Medical Center, Miami, Florida
|| Department of Pediatrics, Division of Neonatology, Christiana Care Health Services, Newark, Delaware
¶ State of Delaware, Public Health Laboratory, Smyrna, Delaware
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Objective. Although common in preterm infants, transient hypothyroxinemia (TH) has not been investigated extensively in ill term infants. The objectives of this study were to investigate serum thyroxine (T4) and thyroid-stimulating hormone (TSH) in sick term infants and to determine whether there is any association between measures of thyroid function and short-term outcome in term infants who receive mechanical ventilation.
Methods. The investigation consisted of both a prospective observational study and a retrospective cohort study. In the prospective study, T4 and TSH were measured after birth in a group of sick term infants (n = 38) and compared with a group of well term infants (n = 18). Infants in the sick group received mechanical ventilation or continuous positive airway pressure and/or had neonatal seizures. Illness severity was quantified using the Score for Neonatal Acute Physiology. The retrospective cohort study included term infants who required mechanical ventilation and were born over a 5-year period (n = 347). Routine T4 screening was collected on the fifth day of life. TH was diagnosed in infants with a T4 <10%, with a TSH <25 µIU/mL. Clinical outcomes in infants with TH were compared with infants without TH.
Results. In the prospective study, infants in the sick group had lower T4 on the fifth day of life as compared with infants in the well group (11.7 ± 4.9 vs 18.9 ± 5.4 µg/dL), and 34% of infants in the sick group had a T4 <10th percentile compared with 6% of infants in the well group. T4 on day of life 5 was inversely correlated with Score for Neonatal Acute Physiology (R = –0.52). In the retrospective study, 21% of mechanically ventilated infants developed TH and were given statistically more inhaled nitric oxide, high-frequency ventilation, vasopressors, and pharmacologic paralysis when compared with infants without TH. Moreover, infants with TH were statistically more likely to die or require transfer to an extracorporeal membrane oxygenation center compared with infants without TH.
Conclusion. Our data show that, similar to preterm infants, ill term infants develop TH. Term infants with TH required more intensive rescue interventions, including inhaled nitric oxide and transfer to an extracorporeal membrane oxygenation center. However, whether T4 levels are a marker or a mediator of clinical outcome remains to be determined.
Key Words: thyroid • illness severity • newborn • term gestation • score for neonatal acute physiology
Abbreviations: TH, transient hypothyroxinemia • T4, thyroxine • TSH, thyroid-stimulating hormone • ECMO, extracorporeal membrane oxygenation • SNAP, Score for Neonatal Acute Physiology • PaO2, arterial oxygen pressure
Transient hypothyroxinemia (TH), characterized by low levels of serum thyroxine and normal levels of thyroid-stimulating hormone (TSH), is a common finding among premature infants.1,2 TH has also been associated with a number of poor outcomes in preterm infants, including death, intraventricular hemorrhage, periventricular leukomalacia, and cerebral palsy.1,3,4 T4 levels in premature infants have also been shown to be inversely correlated with illness severity.5,6 Despite the frequent association of low T4 levels and poor outcome in preterm infants, hypothyroxinemia in term infants has been investigated less extensively. Term infants with perinatal asphyxia and low 5-minute Apgar scores have been shown to have a reduction in thyroid hormone levels after birth.7–9 Similarly, infants who undergo cardiac surgery have been shown to develop euthyroid sick syndrome.10 To our knowledge, hypothyroxinemia in term infants who require mechanical ventilation has not been investigated. This study was designed to answer 2 questions: do sick term infants develop hypothyroxinemia, and is hypothyroxinemia in sick term infants who require mechanical ventilation associated with any differences in short-term outcome?
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
This investigation consisted of 2 components: a prospective observational study and a retrospective cohort study. In both parts of the study, infants were considered term when they were born at 37 weeks’ gestation or greater. All infants were admitted to the observational nursery or the NICU at Christiana Hospital, a regional level III intensive care nursery that cares for both inborn (90%) and outborn infants. Inhaled nitric oxide and high-frequency ventilation both are offered in the NICU at Christiana Hospital. Christiana Hospital, however, does not offer extracorporeal membrane oxygenation (ECMO). Infants who require ECMO are transferred to another institution. The Institutional Review Board of Christiana Care Health Services approved both parts of the study. Written informed consent was obtained from parents of infants who were enrolled in the prospective study.
Prospective Observational Study
Infants were enrolled between August 1, 2000, and July 31, 2001. Infants were enrolled into 2 groups. The study group, or sick group, consisted of term infants who were admitted to the NICU. Infants were included in this group when they were 
37 weeks’ gestation, required mechanical ventilation or continuous positive airway pressure, and/or had the diagnosis of neonatal seizures. Infants with significant congenital anomalies were excluded from the sick group. The control group consisted of "well" infants who were 37 weeks’ gestation, were admitted to the observational nursery for a period <4 hours, and never required admission to the NICU. Infants who were excluded from the well group included those with significant congenital anomalies and any diagnosis or condition leading to the need for intensive care, including sepsis, presumed sepsis, any need for an indwelling intravenous catheter, hypoglycemia requiring parenteral administration of dextrose, respiratory distress, or any need for supplemental oxygen. All infants who were included in the well group were cared for on the routine well infant floor after the brief period of observation and subsequently discharged to home at the same time as their mother. For quantifying illness severity, the Score for Neonatal Acute Physiology (SNAP) was calculated on infants in each group.11
Infants in both groups had blood for total T4 and TSH obtained simultaneously with the first set of routine laboratory tests, which were drawn at <1 hour of life. Blood was obtained by arterial stick, venous stick, or heel stick as appropriate. Because blood was obtained simultaneously with routine laboratory work, none of the infants had extra sticks performed for study purposes. Blood was generally drawn in the well group to obtain a complete blood count, because of risk factors for bacterial sepsis, or to monitor serum glucose because of risk factors for hypoglycemia.
Follow-up T4 was obtained in both groups on the fifth day of life as part of the State of Delaware Newborn Screening Program. Infants in both groups had a total T4 measured at this time. Infants had a TSH measured on the fifth day of life only when they had a total T4 that was <10th percentile for all infants screened in Delaware on that day. For the purposes of this study, TH was defined as it has been previously for preterm infants1: a T4 that was <10th percentile for that day’s screening with a normal TSH level (<25 µIU/mL). The T4 (mean ± SEM) during the study period that defined the 10% was 9.03 ± 0.2 µg/dL, with a coefficient of variation of 20%. All infants with a T4 that was <10th percentile for that day’s screening also had a repeat specimen drawn at 2 to 4 weeks of age.
Power analysis indicated that 51 patients would be required to show a 5-µg/dL difference in T4 between the well and sick groups with an SD of 5 µg/dL, 80% power, and a significance level of .05. The plan was to enroll an equal number of infants in the well and sick groups. However, because of slow recruitment into the well group, the final analysis included a disproportionate number of infants in the sick group. Additional statistical analysis included one-way analysis of variance, Mann-Whitney U test, 
2, and Pearson correlation. P 
.05 was considered significant. All data are expressed as mean ± SD.
Retrospective Cohort Study
Term infants who were born between January 1, 1997, and January 1, 2002, were included when they received mechanical ventilation. Infants with major congenital anomalies, including hemodynamically significant congenital heart disease, were excluded. Routine thyroid screening was collected on the fifth day of life as part of the State of Delaware Newborn Screening Program, and TH was again defined as a T4 level at or below the 10th percentile for that day’s screening with a normal TSH level (<25 µIU/mL).
The need for vasopressor or inotropic support, pharmacologic paralysis, and exogenous surfactant were determined by the attending neonatologist. Inhaled nitric oxide was administered in infants with an oxygenation index >15, and transfer to an ECMO center occurred for infants with an oxygenation index >25 despite maximal medical therapy. Oxygenation index was calculated in standard manner: fraction of inspired oxygen/arterial oxygen pressure (PaO2) x mean airway pressure. The medical team who cared for the infant was not aware of results of the state newborn thyroid screening.
Clinical outcomes in infants with TH were compared with infants without TH. Statistical analysis included analysis of variance and 2. P .05 was considered significant. All data are expressed as mean ± SD.
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Prospective Observational Study
During the study period, 6733 infants were born at Christiana Hospital. A total of 75 infants met the entry criteria for the sick group. Of these 75 infants, 38 infants were enrolled after parental informed consent. During the study period, 22 infants met criteria for the well group. Of these 22 infants, 18 were enrolled after parental consent. There were no differences in gestational age, birth weight, maternal age, proportion of infants who were small for gestational age, gender, mode of delivery, Apgar scores, race, or occurrence of chorioamnionitis or prolonged rupture of membranes between groups (Table 1). As expected by study design, infants in the sick group had a higher SNAP score on the day of birth when compared with the well group.
|
Infants in the sick group included the following diagnoses: 13 (34%) hyaline membrane disease, 9 (24%) pneumonia, 7 (18%) pneumothorax, 5 (13%) transient tachypnea of the newborn or delay in transition to extrauterine environment, and 4 (11%) seizures. None of the infants had culture-proven sepsis. Of the infants in the sick group, 32 (84%) received mechanical ventilation, 6 (16%) received nasal continuous airway pressure, 16 (42%) received exogenous surfactant replacement, and 5 (13%) received dopamine support. One infant in the sick group died before hospital discharge. None of the infants in the sick or well group had a diagnosis of congenital hypothyroidism.
There were no differences in T4 or TSH between the sick and well groups at the time of birth. However, on the fifth day of life, infants in the sick group had a lower T4 compared with infants in the well group (Table 2). In the 13 patients in the sick group who had a measurement of TSH on the fifth day of life, the mean was 5.3 ± 4.3 µIU/mL. Only 1 infant in the well group had a TSH measured on the fifth day of life. Thus, comparison of TSH at this time point could not be made.
|
On the fifth day of life, 13 (34%) of 38 infants in the sick group had a T4 <10th percentile and thus were classified as having TH, compared with 1 (6%) of 18 infants in the well group (P = .03). All 14 infants with TH had repeat testing of T4 and TSH done, and in all cases T4 normalized over time (data not shown).
In the combined population, neither T4 nor TSH was correlated with birth weight (Table 3). T4 at birth was not correlated with illness severity as quantified by SNAP. However, T4 on the fifth day of life was inversely correlated with SNAP. There was no correlation of SNAP with TSH at birth.
|
Within the sick group, there was no correlation between T4 at birth or on the fifth day of life with the lowest PaO2 measured during the first week of life. TSH measured at birth was correlated with the lowest PaO2 during the first week of life (Table 3). Infants who required dopamine support (n = 5) had higher T4 after birth (20.1 ± 9.5 vs 11.4 ± 5.4 µg/dL; P < .01) but a lower T4 on the fifth day of life (6.2 ± 2.5 vs 12.6 ± 4.6 µg/dL; P < .01) compared with infants in the sick group who did not require dopamine support, respectively.
Retrospective Cohort Study
During the study period, 32410 infants were born at Christiana Hospital. Of these births, 347 term infants who did not have major congenital anomalies and received mechanical ventilation were identified and composed the study sample. Infants in the cohort had the following diagnoses: 131 (38%) hyaline membrane disease, 64 (18%) pneumothorax, 63 (18%) persistent pulmonary hypertension of the newborn, 38 (11%) meconium aspiration, 33 (10%) delay in transition to extrauterine environment or transient tachypnea, 26 (7%) pneumonia, and 6 (2%) culture-proven sepsis. Note that some infants had >1 diagnosis; therefore, the numbers do not add up to 100%.
Seventy-two (21%) infants received a diagnosis of TH. There were no differences in gestational age or birth weight in infants with TH compared with the infants without TH (Table 4). Infants with TH were more likely to have been given inhaled nitric oxide, high-frequency ventilation, vasopressors, and pharmacologic paralysis compared with infants without TH (Table 4). Moreover, infants with TH were more likely to die or have been transferred to an ECMO center compared with infants without TH. Infants with TH also had a longer course of mechanical ventilation and had longer length of hospital stay compared with infants without TH (Table 4).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
The main finding of our investigation is that term infants with perinatal illness develop hypothyroxinemia during the first week of life, a finding similar to that commonly observed in preterm infants.1,2 Importantly, term infants with TH had an associated increase in days of mechanical ventilation, length of hospital stay, and use of rescue interventions, including inhaled nitric oxide and transfer to an ECMO center.
In our prospective observational study, 34% of sick group had TH compared with 6% of the well group. The low levels of T4 seen in our study population may have important clinical implications. In preterm infants, TH has been associated with an increase in mortality as well as cerebral palsy.1,3 In our population of term infants, those with TH required more rescue interventions, including inhaled nitric oxide. These infants also required a longer course of mechanical ventilation and had a longer hospital stay compared with infants without TH. It is known that T4 is important for brain growth and development and that in infants with congenital hypothyroidism, early replacement of thyroid hormones is important for optimizing outcome.12 Despite this relationship, preterm infants with hypothyroxinemia have not been shown to derive long-term benefit from thyroid replacement.13
From our data, we cannot determine whether lower levels of T4 are an important part of the causal pathway in perinatal illness or are secondary to illness. Because our study is able to show only an association between TH and the clinical course in term infants who require mechanical ventilation, we do not advocate thyroid hormone replacement on the basis of our data. It is likely that TH is secondary to illness severity. This is highlighted by the fact that sick term infants in our prospective study had lower levels of T4 on the fifth day of life but not at the time of birth. TSH at birth was directly correlated with the lowest PaO2 measured during the first week. Term infants with perinatal asphyxia have previously been shown to have similar levels of total T4, free T4, and TSH in cord blood but lower levels of these hormones at 18 to 24 hours of age, compared with control infants.9 Infants who required dopamine support in our prospective observational study had a reduction in T4 at 5 days of life, despite that they had higher T4 levels after birth. This finding could be reflective of an increased degree of illness or may be secondary, as dopamine has been shown to inhibit the release of TSH.14 Thus, the hypothyroxinemia seen in our population of term infants not only may be related to illness severity but also may be influenced by clinical management.
A randomized trial in term infants who require mechanical ventilation would be needed to determine whether there are any benefits to thyroid hormone supplementation in this population. Any future trials of thyroid replacement in term infants, however, must be developed carefully, as there are potential risks to overtreatment. Previous studies of antenatal TSH-releasing hormone in preterm infants have shown a detrimental effect on developmental outcome.15,16 In the preterm study by Van Wassenaer et al,13 infants in the older gestation category who received T4 replacement had a worse developmental outcome compared with those who received placebo. Studies in preterm infants have advocated an early measurement of thyroid function, which could lead to a strategy of treating only infants with the lowest T4 rather than treating on the basis of risk factors such as gestational age or the need for mechanical ventilation.17 Treating on the basis of some measurement of thyroid function rather than a risk factor such as mechanical ventilation would likely prove most efficacious as only 21% of mechanically ventilated term infants in our retrospective study had TH. However, future studies in term infants will be needed to determine the low T4 level that warrants replacement therapy. In our study, we defined TH by a T4 <10%, which is based on state newborn screening policy and past studies of TH in preterm infants.1 It is possible, however, that this statistical definition may not reflect the functional lower limit of T4 and therefore may not be the appropriate cutoff level at which to initiate therapy.
Additional potential clinical significance of thyroid function after birth was recently highlighted by Larson et al,18 who showed that both very low birth weight and non–very low birth weight infants who require intensive care are at increased risk for developing congenital or transient hypothyroidism identified by delayed elevation of TSH.18 Larson et al did not study illness severity but were able to associate the use of dopamine and iodine with delayed elevation in thyrotropin.18 None of the infants in either our prospective or retrospective cohort study exhibited delayed elevation in TSH, but this may be because of the number of infants in our study. Studies on larger populations of term infants who require intensive care are required to determine whether the hypothyroxinemia seen in our population of term infants may be a precursor for later disorders of thyroid function, including delayed elevation in thyrotropin and hypothyroidism that requires treatment.
Our study has a number of important limitations. T4 and TSH are known to surge after birth. We chose to measure T4 and TSH at this time to determine whether antenatal conditions affect thyroid function and whether T4 and TSH surge differently in well and sick term infants. We did not measure other important thyroid hormones in the prospective observational study, such as free T4, triiodothyronine, thyroid hormone binding globulin, or reverse triiodothyronine, because of issues of obtaining large amounts of blood in well infants. Although free T4 is the biologically active form of the hormone, total T4 has been studied extensively in the very low birth weight population and has been associated consistently with neonatal morbidity and mortality.1,3,4,19 Moreover, in the preterm population, free T4 has been demonstrated to correlate closely with total T4.19 In term infants, both low free and total T4 have been demonstrated after low Apgar scores.8 Additional studies investigating free T4 with outcome in term infants may help to determine whether TH is causal in the need for increased rescue therapies in infants who require mechanical ventilation. Infants were eligible for the well group, which served as a control, only when they were having blood drawn for another clinical indication. This was done to avoid needle sticks in otherwise healthy infants. Because of our study design, infants in the well group had risk factors for neonatal illness or even a modicum of illness. This study design may have led to an underestimation of the differences between study groups. However, none of the infants in the well group had sufficient illness to require intensive care, and the large difference in SNAP between the sick and well groups confirms that the 2 groups had a large discrepancy in illness severity. Many of the outcomes measured in the retrospective cohort component of our study, such as the need for inhaled nitric oxide or pressor support, are based on clinical management decisions. However, our data are from a single center with uniform clinical management protocols. Furthermore, the clinicians who ordered rescue therapies such as inhaled nitric oxide were unaware of results from newborn screening when making management decision.
In summary, our data show that term infants with perinatal illness are more likely to develop hypothyroxinemia compared with a group of well infants and that infants who require mechanical ventilation and develop TH are more likely to require rescue interventions. Because sick infants are more likely to develop TH than well infants, our data suggest that illness severity must be accounted for when interpreting results from newborn screening within the first week of life. Because additional investigation, including long-term outcome, is needed to determine whether T4 levels are a mediator or simply a marker of clinical status, we do not advocate thyroid hormone replacement on the basis of our study.
|
FOOTNOTES |
|---|
Accepted Jul 7, 2004.
Reprint requests to (D.A.P.) Division of Neonatology, Christiana Care Health Services, 755 Ogletown Stanton Rd, Newark, DE 19178. E-mail: paul.d{at}christianacare.org
No conflict of interest declared.
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
- Paul DA, Leef KH, Stefano JL, Bartoshesky L. Low serum thyroxine on initial newborn screening is associated with intraventricular hemorrhage and death in very low birth weight infants.
Pediatrics. 1998;101
:903
–907
[Abstract/Free Full Text] - Frank JE, Faix JE, Hermos RJ, et al. Thyroid function in very low birth weight infants: effects on neonatal hypothyroidism screening. J Pediatr. 1996;128 :548 –554[CrossRef][Web of Science][Medline]
- 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
–827
[Abstract/Free Full Text] - Leviton A, Paneth N, Reuss ML, et al. Hypothyroxinemia of prematurity and the risk of cerebral white matter damage. J Pediatr. 1999;134 :706 –711[CrossRef][Web of Science][Medline]
- Paul DA, Leef KH, Voss B, Stefano JL, Bartoshesky L. Thyroxine and illness severity in very low-birth-weight infants. Thyroid. 2001;11 :871 –875[CrossRef][Web of Science][Medline]
- van Wassenaer AG, Kok JH, Dekker FW, de Vijlder JJ. Thyroid function in very preterm infants: influences of gestational age and disease. Pediatr Res. 1997;42 :604 –609[Web of Science][Medline]
- Tahirovic HF. Transient hypothyroxinemia in neonates with birth asphyxia delivered by emergency cesarean section. J Pediatr Endocrinol. 1994;7 :39 –41[Web of Science][Medline]
- Sak E, Akin M, Akturk Z, et al. Investigation of the relationship between low Apgar scores and early neonatal thyroid function. Pediatr Int. 2000;42 :514 –516[CrossRef][Web of Science][Medline]
- Pereira DN, Procianoy RS. Effect of perinatal asphyxia on thyroid-stimulating hormone and thyroid hormone levels. Acta Paediatr. 2003;92 :339 –345[CrossRef][Web of Science][Medline]
- Allen DB, Dietrich KA, Zimmerman JJ. Thyroid hormone metabolism and level of illness severity in pediatric cardiac surgery patients. J Pediatr. 1989;114 :59 –62[CrossRef][Web of Science][Medline]
- Richardson DK, Gray JE, McCormick MC, Workman K, Goldmann DA. Score for Neonatal Acute Physiology: a physiologic severity index for neonatal intensive care.
Pediatrics. 1993;91
:617
–623
[Abstract/Free Full Text] - Timiras PS, Nzekwe EU. Thyroid hormones and nervous system development. Biol Neonate. 1989;55 :376 –385[Web of Science][Medline]
- van Wassenaer AG, Kok JH, de Vijlder JJ, et al. Effects of thyroxine supplementation on neurologic development in infants born at less than 30 weeks’ gestation.
N Engl J Med. 1997;336
:21
–26
[Abstract/Free Full Text] - Seri I, Somogyvari Z, Hovanyovszky S, Kiszel J, Tulassay T. Developmental regulation of the inhibitory effect of dopamine on prolactin release in the preterm neonate. Biol Neonate. 1998;73 :137 –144[CrossRef][Web of Science][Medline]
- Crowther CA, Hiller JE, Haslam RR, Robinson JS. Australian Collaborative Trial of Antenatal Thyrotropin-Releasing Hormone: adverse effects at 12-month follow-up. ACTOBAT Study Group.
Pediatrics. 1997;99
:311
–317
[Abstract/Free Full Text] - Briet JM, van Sonderen L, Buimer M, Boer K, Kok JH. Neurodevelopmental outcome of children treated with antenatal thyrotropin-releasing hormone.
Pediatrics. 2002;110
:249
–253
[Abstract/Free Full Text] - Kantor MJ, Leef KH, Bartoshesky L, Getchell J, Paul DA. Admission thyroid evaluation in very-low-birth-weight infants: association with death and severe intraventricular hemorrhage. Thyroid. 2003;13 :965 –969[CrossRef][Web of Science][Medline]
- 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 –591[CrossRef][Web of Science][Medline]
- Biswas S, Buffery J, Enoch H, Bland JM, Walters D, Markiewicz M. A longitudinal assessment of thyroid hormone concentrations in preterm infants younger than 30 weeks’ gestation during the first 2 weeks of life and their relationship to outcome.
Pediatrics. 2002;109
:222
–227
[Abstract/Free Full Text]
PEDIATRICS (ISSN 1098-4275). ©2005 by the American Academy of Pediatrics
CiteULike 
Connotea 
Del.icio.us 
Digg 
Facebook 
Reddit 
Technorati 
Twitter What's this?
This article has been cited by other articles:
|
|
 |
|
  M. J. E. Kempers, C. I. Lanting, A. F. J. van Heijst, A. S. P. van Trotsenburg, B. M. Wiedijk, J. J. M. de Vijlder, and T. Vulsma Neonatal Screening for Congenital Hypothyroidism Based on Thyroxine, Thyrotropin, and Thyroxine-Binding Globulin Measurement: Potentials and Pitfalls J. Clin. Endocrinol. Metab., September 1, 2006; 91(9): 3370 - 3376. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
