Transient neonatal hyperthyrotropinaemia

Early Human Development, 28 (1992) 19-25 Elsevier Scientific Publishers Ireland Ltd.

19

EHD 01213

Transient neonatal hyperthyrotropinaemia Terence

T. Laoa, Chi Yin Li” and Nirmal

S. Panesarb

ODepartment of Obstetrics and Gynaecology and bDepartment of Chemical Pathology, Prince of Wales Hospital, Shatin (Hong Kong) (Received

15 May 1991; revision

received

2 November

1991; accepted

5 November

1991)

Summary

Of 48 consecutive newborns with elevated umbilical venous plasma thyrotropin (TSH) concentration, only two (4%) were subsequently proved to have congenital hypothyroidism, while the other 46 had transient elevation of TSH. Compared with matched controls, these 46 newborns were all delivered vaginally (P < 0.0003) and had a longer second stage of labour (P < 0.002), together with higher incidences of nuchal encirclement of the cord (P < 0.05) as well as female babies (P < 0.05). There was no difference in the incidence of antenatal complications, mean gestational age, birth weight, or birth asphyxia. There were no small-for-gestational age infants in the study group, while four were found in the controls. The results indicate that elevated umbilical cord plasma TSH concentration may represent a response to the stress of difficult or complicated delivery in the healthy appropriate- or large-forgestational age newborn who does not have congenital hypothyroidism. Key words: thyrotropin;

enthyroid neonates; parturition

Introduction

The measurement of umbilical cord plasma thyrotropin (TSH) concentration is one of the established screening tests for congenital hypothyroidism [ 11. Occasional cases of unexplained elevated TSH concentration have been found in neonates subsequently shown to be not hypothyroid [I ,2]. However, the underlying cause for this phenomenon is uncertain. The results of our earlier study [3] and that of Fukuda (1987) [4] suggest an association between umbilical cord plasma TSH concentration Correspondence to: T.T. Lao, Perinatal Toronto, Ont. MSG 1X5, Canada.

Unit, Suite 775, Mount

0378-3782/92/$05.00 0 1992 Ekevier Printed and Published in Ireland

Scientific

Publishers

Sinai Hospital,

Ireland

Ltd.

600 University

Avenue,

20

and the mode of delivery. In view of this finding, we have studied another 48 consecutive newborns with elevated TSH (> 20 m1.U.Q and compared them with matched controls to determine the factors, apart from congenital hypothyroidism, that may be associated with elevated umbilical cord serum TSH. Patients and Methods

In our hospital, most of the normal newborns are discharged from hospital with their mothers within 3 days of birth. Thus to screen for congenital hypothyroidism, a S-ml sample of umbilical venous blood is collected into a heparinised tube at birth for the assay of TSH and thyroxine in all newborns. The samples are stored at 4°C until separation of the plasma, and TSH assay using a commercial radioimmunoassay kit (Diagnostic Products Corporation) is performed on the daily batches. Neonates with an elevated TSH are recalled for a repeat TSH measurement between 1 and 2 weeks after birth. Forty-eight consecutive newborns with elevated umbilical venous plasma TSH (upper limit of normal 20 mI.U./l) delivered over a 3-month period and represented 2.6% of the samples collected were identified from the laboratory records in a retrospective survey. These newborns were all from singleton pregnancies. For each study case, a list of singleton newborns delivered on the same day with normal umbilical venous TSH concentration (< 20 mI.U./l) tested in the same batch was identified, and the first one on the list whose maternal age and parity, as indicated on the delivery suite patient registry, matched the study case was selected as control. This was done by one of the authors who was blinded to the outcome of pregnancy. In all controls, the parity was exactly matched but in seven controls, the maternal age was f 2 years from the respective study case because an exact match was not available. The match for the age and parity was further confirmed from the hospital notes of the study and control cases which were retrieved for analysis. The antenatal and intrapartum complications, type of labour, mode of delivery and outcome were compared between these two groups. Paired and unpaired t-test, Mann Whitney Utest, x2-test and Fisher’s exact probability test were used for statistical analysis where appropriate.

All the 48 newborns in the study group and their controls were from singleton pregnancies. All newborns in the study group had TSH concentration above 20 mI.U./l, and 31.3% (15148) and 10.4% (5148) had initial TSH concentrations above 40 mI.U./l and 50 mI.U./l, respectively. The figures were adjusted to 30.4% (14/46) and 8.7% (4/46) after excluding two cases of congenital hypothyroidism which were confirmed by the repeated TSH assay and subsequent clinical course. Their initial TSH results were 158 mI.U./l and 25.8 mI.U./l, respectively. The initial and repeated TSH concentrations in the remaining 46 study cases are shown in Fig. 1. The mean age at follow-up TSH estimation was 8 days. The matched controls for the two mother-newborn pairs with congenital hypothyroidism were also excluded from subsequent analyses.

21

100 I_

90l80, 70 I 60 I

_

50,

_

40

30

20

d .i 10 59 P p

0 7

Ofi 5 4

3

2

1

AT

Fig.

1. Postnatal

decrease

BIRTH

in plasma

FOLLOW UP

thyrotropin

concentrations

in hyperthyrotropinaemic

euthyroid

neonates.

There were 34 (73.9%) nulliparous mothers each in the study and control groups, and the parity in the rest ranged between one and three, with a mean parity of 1.6. The maternal age (mean f SD) in the study and control groups were 27.7 + 3.9 years and 27.6 f 3.5 years, respectively. The antenatal complications in both groups are shown in Table I. The overall incidence of antenatal complications were similar (19.6% in the study group vs. 21.7% in the control group). All the cases of anteparturn haemorrhage were of unknown origin, occurring between 27 and 32 weeks.

22 TABLE

I

Antenatal

complications

in the study and control

groups

(“h).

Study group (n = 46)

Control group (n = 46)

Pre-eclampsia Asthma

2 (4.3%) 0

0 2 (4.3%)

Anaemia Antepartum haemorrhage Pneumatic heart disease Gestational diabetes

3 (6.5%) 0 0 2 (4.3”/;,)

0 4 (8.7”/;,) 1 (2.2%) 0

Schizophrenia Prolonged rupture

1 (2.2%) 1 (2.2%)

0 3 (6.5%)

of membranes

None of the mothers in either groups had thyroid disease, antithyroid treatment, or thyroxine replacement. There were no differences in the incidence of induced labour or epidural analgesia between the groups, but all the mothers in the study group had undergone labour (Fisher’s test P < 0.03) and none were delivered by caesarean section (Fisher’s test P < 0.0003) (Table II). For mothers with vaginal delivery, there was no difference in the mean duration of the first stage of labour, but the mothers in the study group had a significantly longer second stage (Mann Whitney U-test P < 0.002) and a

TABLE

II

Labour

and delivery

in the study and control

groups. Study group (n = 46)

Type of labour No labour

0"

Control group (n = 46)

5 (10.9%)

Induction Spontaneous Epidural analgesia Mode of delivery

7 (15.2%) 39 (84.8%) 2 (4.3%)

5 (10.9%) 36 (78.3%) 2 (4.3%)

Spontaneous vaginal Instrumental vaginal Caesarean section Duration of labour in

34 (73.9%) 12 (26.1%) Ob

30 (65.2%) 5 (10.9%) I I (23.9%)

n = 321 37 8

n = 289 26 3

vaginal delivery groups First stage (mean f S.E.M.) min second stage (mean f S.E.M.) min Number with prolonged second stage Spearman’s Correlation between first and second

46 f 28 f 3c (I 7.4%)

-0.009 stages

“Fisher’s test P = 0.0279. bFisher’s test P = 0.00025. CMann Whitney U-test P = 0.001 I. dSpearman’s Rank Correlation P < 0.05.

35 f 40 f 4 (6.5%)

0.363d

23 TABLE Outcome

III of labour

and delivery

in study and control

groups

(%).

Study group (n = 46) Complications Meconium stained liquor Nuchal encirclement of the cord Shoulder dystocia Postpartum haemorrhage Fetal outcome Gestation (mean f S.D.) weeks number < 37 weeks number > 42 weeks Birth weight (mean f S.D.) g number < 2500 g number > 4000 g Apgar score (1 min) < 7 Sex Male Female

11 6 1 3

(23.9%) (13.0%)” (2.2%) (6.5%)

39.0 f

Control group (n = 46)

6 (13.0%) 1 (2.2%) 0 1 (2.2%)

1.3

7 L

1 3190 f 416 0 3 2 (4.3%) 17b 29

28 18

aFisher’s test P < 0.05. bx2-test P < 0.05.

higher, though statistically not significant, incidence of prolonged second stage (> 60 min in the nulliparous and > 30 min in the multiparous mother). When the relationship between the first and second stages was analysed, a significant correlation (Spearman’s Rank Correlation = 0.363, P c 0.05) was found only in the controls. When the outcome of labour and delivery was analysed, significant meconiumstained liquor (moderate to thick) was found more often in the study group, though the difference just failed to reach statistical significance (Table III). The study group however had a significantly higher incidence of nuchal encirclement of the cord There was no difference in the mean gestation at delivery (Fisher’s test P < 0.05). or birth weight, or in the number of premature and postdated deliveries. In the study group three babies had birth weight > 4000 g, and six were large-for-gestational age (> 90th centile) but none were small-for-gestational age (< 10th centile). In the control group, two babies had birth weight < 2500 g, with four small-for-gestational age and three large-for-gestational age babies. However, these differences were not statistically significant. There were more female babies born in the study group compared to the control group @-test P < 0.05). When the 14 cases (30.4%) in the study group with TSH concentration > 40 mI.U./l were further analysed, they were not associated with a significantly higher proportion of complications (prolonged second stage 37.5%, instrumental delivery 16.7%, nuchal encirclement of the cord 42.9%, meconium staining of the liquor 27.3%).

24

Discussion

An increased umbilical cord blood cortisol concentration is found in infants born vaginally compared to infants born by caesarean section [5,6]. This phenomenon is also observed in preterm infants [7]. This is thought to represent the fetal response to the stress of intrauterine hypoxia and acidosis during labour [8], and is probably mediated by the pituitary release of adrenocorticotropic hormone (ACTH) which is also found to be significantly elevated in cord arterial blood after vaginal delivery versus caesarean section [9]. Thus the hypothalamic-pituitary axis probably plays an important role in the fetal response to intrapartum stress. Cord arterial TSH concentration is significantly increased following vaginal delivery compared to elective caesarean section [4]. This appears to be related to the route of delivery rather than the presence of labour, as lOoh of non-asphyxiated singleton full term newborns delivered vaginally had cord blood TSH concentration above 20 mI.U./l while none is found in the newborns delivered by caesarean section [3]. Thus the postnatal thyrotropin surge, which reaches a peak 30 min after birth [lo], is probably another manifestation of the hypothalamic-pituitary response to the stress of vaginal delivery in addition to the response of the fetus to exposure to extrauterine environment, since infants born by caesarean section have higher thyrotropin concentration on the fifth postpartum day [ 11,121. Therefore an increase in thyrotropin concentration, also found in otherwise healthy women undergoing elective major surgical procedures for benign gynaecological conditions [13], is likely to represent part of the normal hypothalamic-pituitary response to stress in the healthy human. In our study cases, the type of labour was probably dystocic, since the first stage was slightly, and the second stage significantly, longer than controls, with a 3-fold increase in the incidence of prolonged second stage. The lack of correlation between the first and second stages of labour suggests that the dystocia was related to the passage, since the birth weight was similar to the controls and there was no difference in the incidence of epidural analgesia. As well, the significantly higher incidence of nuchal encirclement of the cord in the study group could be additional or in some cases, the primary cause of stress. One newborn (birth weight 3800 g) with shoulder dystocia was found in the study group. The second stage was 60 min and initial TSH was 52.2 mI.U./l, but the Apgar scores were 8 and 10 at 1 and 5 min, respectively. On the other hand, none of the newborns in the study group had severe birth asphyxia (Apgar score < 4), intrauterine growth retardation or congenital anomalies. These observations suggest that the peripartum surge in fetal TSH in association with difficult or complicated vaginal delivery is a response found only in normal healthy appropriate- or large-for-gestational age fetuses. The finding that newborns with severe neonatal diseases (including asphyxia and neonatal infections) tended to have normal TSH concentrations [14] supports this hypothesis. Our results suggest that transient neonatal hyperthyrotropinaemia represents a fetal response to the acute hypoxic stress of vaginal delivery, especially complicated delivery, such as described in a case by Walfish (1976) [2]. This response may be mediated by noradrenaline, which was significantly correlated with the TSH concentration [4]. Female babies exhibit such a response more frequently. Since the newborns with very high TSH concentration (> 40 mI.U.11) accounted for a propor-

25

tionately similar incidence of the various peripartum complications in the study group, this TSH surge appears to be a qualitative rather than quantitative response to parturition stress in otherwise normal infants. Our finding indicates that screening of congenital hypothyroidism using cord blood TSH concentration can be associated with a high false positive rate. In view of this, and when circumstances permit, the use of the fifth postpartum day sample should be the preferred screening test. Acknowledgement

The authors are grateful to Dr Y. M. Lam, PhD, for his help and advice in statistical analysis. References 1 2 3 4 5

6 7 8 9 10 II 12

13 14

John, R. (1987): Screening for congenital hypothyroidism. Ann. Clin. Biochem., 24, l- 12. Walfish, P.G. (1976): Evaluation of three thyroid function screening tests for detecting neonatal hypothyroidism. Lancet, i, 1208-1211. Lao, T.T. and Panesar, N.S. (1989): Newborn thyrotropin and mode of delivery. Br. J. Obstet. Gynaecol., 96, 1224-1227. Fukuda, S. (1987): Correlation between function of the pituitary-thyroid axis and metabolism of catecholamines by the fetus at delivery. Clin. Endocrinol., 27, 331-338. Norman, R.J., Deppe, W.M., Coutts, P.C., Marivate, M. and Joubert, SM. (1983): Twin pregnancy as a model for studies in fetal cortisol concentrations in labour: relation to prostaglandins, prolactin and ACTH. Br. J. Obstet. Gynaecol., 90, 1033-1039. Lao, T.T. and Panesar, N.S. (1989): The effect of labour on prolactin and cortisol concentrations in the mother and fetus. Eur. J. Obstet., Gynecol. Reprod. Biol., 30, 233-238. Procianoy, R.S., Cecin, S.K.G. and Pinheiro, C.E.A. (1983): Umbilical cord cortisol and prolactin levels in preterm infants. Acta Paediatr. Stand., 72, 713-716. Martinsen, K., Peltoia, J., Tervila, L. and Virtanen, A. (1982): Umbilical cord cortisol and arterial pH levels in spontaneous and induced labours. Obstet. Gynecol., 59, 171-175. Puolakka, J., Kauppila, A., Tuimala, R. and Pakarinen, A. (1982): Fetal adrenocorticotropic mone and prolactin at delivery. Obstet. Gynecol., 60, 71-73. Fisher, D.A. and Klein, A.H. (1981): Thyroid development and disorders of thyroid function newborn. N. Engl. J. Med., 304, 702-712.

horin the

Engberg, S., Gustavson, K.H., Jacobsson, L. and Soederstroem, R. (1978): Elevation of TSH during the early neonatal period. J. Pediatr., 92, 1030. Larsson, A., Ljunggren, J.G., Ekman, K., Nilsson, A., Olin, P. and Bodegard, G. (1981): Screening for congenital hypothyroidism. 11. Clinical findings in infants with positive screening tests. Acta Paediatr. Stand., 70, 147-153. Adashi, E.Y., Rebar, R.W., Ehara, Y., Naftolin, F. and Yen, S.S.C. (1980): Impact of acute surgical stress on anterior pituitary function in female subjects. Am. J. Obstet. Gynecol., 138, 609-614. Klett, M., Bohnert, R., Schoenberg, D. and Wille, L. (1981): Thyroid function in newborns suffering from severe neonatal

diseases.

Acta Endocrinol.

Suppl.,

240, 8-9.