ORIGINAL ARTICLE

E n d o c r i n e C a r e

Screening for Congenital Hypothyroidism: The

Significance of Threshold Limit in False-Negative
Results
Chryssanthi Mengreli, Christina Kanaka-Gantenbein, Panagiotis Girginoudis,
Maria-Alexandra Magiakou, Ioulia Christakopoulou, Aglaia Giannoulia-Karantana,
George P. Chrousos, and Catherine Dacou-Voutetakis
Department of Biochemical Laboratories (C.M., P.G., A.G.-K.), Institute of Child Health, Aghia Sophia
Children’s Hospital, Division of Endocrinology, Metabolism, and Diabetes (C.K.-G., M.-A.M., G.P.C.,
C.D.-V.), First Department of Pediatrics, Athens University Medical School, and Department of Nuclear
Medicine (I.C.), Sotiria Hospital, Athens 11527, Greece

Context: In our neonatal program, a number of infants with congenital hypothyroidism (CH) had
escaped diagnosis, when a spot RIA-TSH value of 20 mU/liter whole blood was used as a cutoff point.

Objective: The objective of the study was to find out prospectively the additional number of
newborns with CH if the TSH cutoff point is lowered to 10 mU/liter.

Population and Methods: The study included 311,390 screened newborns. The children with CH
were followed up for a period of 3 yr.

Results: Twenty-eight percent of infants diagnosed with CH had neonatal TSH values between 10
and 20 mU/liter (56 of 200). Forty of 47 infants, who were reevaluated later on (85.1%), suffered
permanent CH. A thyroid scintiscan and/or echogram revealed that eight of 40 children (20.0%) had
a structural defect, and the remaining (32 of 40) had a functional defect of the thyroid gland
without anatomical abnormality; 14 of 32 cases were familial. Eighteen of the 47 reevaluated
infants were prematurely born (38.3%) and 15 of these 18 had permanent CH (83.3%). The low-
ering of TSH cutoff point from 20 to 10 mU/liter resulted in a 10-fold increase of recall rate.

Conclusions: A significant number of cases with permanent CH are missed when a TSH threshold
of 20 mU/liter is applied. Almost 40% of the missed CH cases were premature. A mild increase of
TSH at screening is not a predictor of transient CH. The increase in recall rate constitutes a serious
drawback and should be balanced against the possible consequences of thyroid dysfunction at this
important developmental stage. (J Clin Endocrinol Metab 95: 4283– 4290, 2010)

creening programs for congenital hypothyroidism and particularly TSH threshold have not as yet satisfac-
S (CH) have virtually eradicated mental retardation
and impaired somatic growth caused by thyroid hormone
torily been defined. This was obvious in a survey, which
outlined the status of neonatal screening in Europe in 2004
deficiency. This has been achieved by early diagnosis and (3). Screening TSH cutoff values ranging between 5 and 25
treatment with L-thyroxine (1). Hence, screening pro- mU/liter were applied in 178 screening centers, using a
grams are cost effective to society, a fact that is taken into variety of methods for TSH measurement and showed dif-
consideration by governmental health officials (2). ferences in the prevalence of CH and in recall rates.
Although the screening programs for CH were estab- Numerous studies reported missed CH cases during the
lished more than 30 yr ago, the optimal screening strategy operation of neonatal screening programs. The causes of

ISSN Print 0021-972X ISSN Online 1945-7197 Abbreviations: BW, Birth weight; CH, congenital hypothyroidism; FT4, free T4.
Printed in U.S.A.
Copyright © 2010 by The Endocrine Society
doi: 10.1210/jc.2010-0057 Received January 11, 2010. Accepted May 21, 2010.
First Published Online June 30, 2010

For editorial see page 4223

J Clin Endocrinol Metab, September 2010, 95(9):4283– 4290 jcem.endojournals.org 4283

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 07 June 2015. at 13:34 For personal use only. No other uses without permission. . All rights reserved.
4284 Mengreli et al. Screening for Congenital Hypothyroidism J Clin Endocrinol Metab, September 2010, 95(9):4283– 4290

CH in these missed cases were either defects in T4 synthesis imen on a new Guthrie card was asked by mail. On this repeat
(4 – 6) or thyroid gland malformations (7). specimen, TSH values less than 8 mU/liter were considered nor-
mal, taking into account that the infant was usually more than 1
The newborn screening for CH was initiated in Greece
month old by the time of reexamination. A TSH result greater
as a pilot study in 1979 (8) and was gradually generalized than 20 mU/liter on the initial Guthrie card was considered pos-
and covered the total newborn population by 1990. A itive for CH and the newborn was referred for biochemical and
blood spot TSH value of 30 mU/liter was initially used as clinical evaluation immediately. The same was also applied when
cutoff point, which was lowered to 20 mU/liter in the TSH concentration on the repeat Guthrie card was greater than
1990s, a value recommended by the Working Group of the 8 mU/liter and the infant older than 30 d.
CH was confirmed and replacement therapy with L-thyroxine
European Society of Pediatric Endocrinology (9). Based
was initiated when serum TSH levels were greater than 10 mU/
on this criterion, an overall prevalence of CH of 1:3300 liter and T4 or free T4 (FT4) levels were below or low-normal for
screened newborns was found (10), a figure comparable the age-related reference range. When TSH serum levels were
with that in other European countries (11) and the United between 5 and 9 mU/liter (gray zone) and T4 or FT4 levels were
States (12). During the screening operation, however, it normal, the baby was followed up every 15 d. Infants with re-
peatedly increased serum TSH levels (⬎5 mU/liter) up to 2–3
became evident that there were some CH cases, which had
months of life also received a substitution therapy with
escaped diagnosis, because they presented with spot TSH L-thyroxine.
values below the cutoff point applied by that time (false
negative results). These patients were identified by their Methods
pediatricians, who suspected CH either on the basis of
clinical symptoms and/or signs of CH or because they or- Methods at screening level
dered thyroid function tests due to familial CH. This ob- An in-house RIA-TSH method was applied in the program.
The reagents were produced by the Institute of Radioisotopes
servation prompted the initiation of a prospective study in
and Diagnostic Products of the Research Center Dimokritos
which a TSH threshold of 10 mU/liter was used, aiming at (Greece). The calibration of TSH standards was done against the
detecting the number of false-negative cases and their un- second International Standard, National Institute for Biological
derlying pathogenetic defect. Standards and Control code 80/558. The sensitivity of the
method was 5 mU/liter. The results are expressed as milliunits per
liter whole blood (milliunits per liter whole blood ⫻ 2.2 ⫽ mil-
liunits per liter serum, assuming an hematocrit of 55% for the
Subjects and Methods first days of life). Internal and external quality assurance pro-
grams were included in the TSH analysis.
Subjects
In Greece there is one screening center located at the Institute Assay performance
of Child Health in Athens, which covers the newborns of the Interassay coefficients of variation for blood spot TSH assays
whole country (about 10,500,000 inhabitants). More than were 10.9 and 12% for the RIA method at two TSH levels,
100,000 neonates born in about 160 state and private maternity namely 12 and 26 mU/liter, respectively.
units all over the country are screened every year by measuring
blood spot TSH concentrations. Blood is taken by heel prick
between the third and fifth day of life when the baby is discharged Methods at diagnosis of CH
from the hospital, and the Guthrie cards are mailed to the Insti- The diagnosis of CH was confirmed by measurements of se-
tute of Child Health daily or every second day. Premature or sick rum T4 and/or FT4, T3, and TSH. Complementary serum anti-
full-term neonates, who usually have a prolonged stay in neo- thyroglobulin antibodies, antiperoxidase antibodies, antibodies
natal intensive care units, are screened twice, once between the against TSH receptor, and thyroglobulin as well as urine iodine
fifth and seventh day of life, irrespective of feeding status and a were measured for additional information concerning the cause
repeat time on discharge from the neonatal intensive care unit. In of the disease and its permanent or transient character. The
cases of prolonged hospitalization, a second Guthrie card may be above parameters (serum T4, T3, FT4, TSH, and antithyroid
discretionary sent by 2–3 wk of age. antibodies) were also measured in the mothers of the investigated
The study was prospective and had been carried out between infants.
January 2000 and December 2002. During this 3-yr period, Commercially available reagents from Roche Diagnostics
311,390 newborns were screened. (Mannheim, Germany) were used for the determinations of se-
rum T4, T3, FT4, TSH, antithyroglobulin antibodies, antiper-
oxidase antibodies, and thyroglobulin. Analysis was carried out
Screening protocol of the study on an Elecsys 2010 apparatus (Hitachi, Tokyo, Japan), and it
When a TSH value on Guthrie cards was lower than 10 mU/ was based on an electrochemical chemiluminescence technique.
liter whole blood, it was considered negative and no further The antibodies against TSH receptor were measured by an
action was pursued. When it was higher than 10 mU/liter, a immunoradiometric method using Brahms (Hennigsdorf, Ger-
second measurement on blood from the initial Guthrie card was many) reagents, which detect immunoreactive TSH binding (ei-
performed in duplicate. If the repeated measurements were less ther inhibiting or stimulating) immunoglobulins.
than 10 mU/liter they were considered negative. Results between Urinary Iodine was assessed by a colorimetric method of Pino
10 and 20 mU/liter were designated borderline and a blood spec- et al. (13) in a spot urine sample and was expressed as micro-

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 07 June 2015. at 13:34 For personal use only. No other uses without permission. . All rights reserved.
J Clin Endocrinol Metab, September 2010, 95(9):4283– 4290 jcem.endojournals.org 4285

grams per liter of urine. Median urinary iodine concentrations in

Group A, Infants with neonatal TSH values between 10 and 20 mU/liter whole blood; group B, infants with neonatal TSH values greater than 20 mU/liter whole blood. Infants ⬙without treatment⬙ are
those who did not receive L-thyroxine therapy because serum TSH values at the confirmatory examination were less than 5 mU/liter (false positives). Conversion factor: neonatal TSH milliunits per liter
0.000
0.000
0.000
0.000
groups of at least 30 children less than 2 yr old of greater than 100

0.05
ns
ns

ns
P
␮g/liter are indicating adequate iodine intake, whereas values of
less than 100 ␮g/liter insufficient iodine intake (14).

Methods during replacement therapy

Without treatment

3040 ⫾ 668.4 (69)c

25.7 ⫾ 21.2 (69)b

183.1 ⫾ 53.2 (52)d

9.65 ⫾ 2.14 (63)c
12.9 ⫾ 2.54 (28)c

3.8 ⫾ 2.0 (69)d

Ultrasonography of the thyroid gland was performed during

11 (15.9)a
(n ⴝ 70)
the first months of life in two centers.

125 (20)d
Group B (n ⴝ 194)
Methods at reevaluation
First, by 3 yr of age and after a 3-wk withdrawal of L-thy-
roxine, a thyroid scintiscan with I123 was carried out followed by
a perchlorate discharge test if a thyroid gland orthotopic normal
in size and shape had been found on ultrasound. A radioactive
iodine discharge of more than 90% of the basal uptake measured

140.5 ⫾ 81.3 (104)b

21.3 ⫾ 10.5 (124)a

4.7 ⫾ 3.78 (113)a

3015 ⫾ 738 (123)b

365 ⫾ 396 (125)a

Under treatment
half an hour after sodium perchlorate administration indicates

5.46 ⫾ 5.4 (99)a

total iodide organification defect, whereas discharge values be-

(n ⴝ 124)

22 (17.9)a

370 (97)d
tween 10 and 90% are suggestive of partial iodide organification
defects. All the isotope scans were interpreted by the same op-
erator. Second, if the infants were well controlled (TSH levels in

Various parameters of recalled infants who were examined for confirmation of diagnosis of CH
the normal range) with a relatively low L-thyroxine dose, thyroid
hormone replacement was discontinued earlier than 3 yr,
namely, between 12 and 24 months of age.
The diagnosis of permanent CH was based on: 1) a serum

0.001
TSH concentration greater than 5 mU/liter after discontinuation

ns
ns
ns
ns
ns
ns

ns
P
of therapy for at least 3 wk, 2) the presence of siblings with an
increased serum TSH, and 3) abnormal image studies or func-
tional tests.
55.71 ⫾ 38.06 (66)b
2791.3 ⫾ 821.9 (66)c
Without treatment

Reference ranges for full-term infants for T4, FT4, T3, and

174.5 ⫾ 59.8 (56)d

4.03 ⫾ 1.65 (66)d
11.59 ⫾ 2.38 (32)c
8.83 ⫾ 2.3 (62)c
TSH are 6.5–16.3 ␮g/dl (84 –210 nmol/liter), 9 –26 pg/ml (12–33
pmol/liter), 100 –300 ng/dl (1.5– 4.5 nmol/liter), and 0.8 –5 mU/ 20 (30.3)a
(n ⴝ 66)
Group A (n ⴝ 120)

65 (6)d
liter, respectively. Reference ranges for 1- to 4-wk-old preterm
newborns for T4, FT4, T3, and TSH were adopted by the study

P ⬍ 0.0001, comparisons between infants of groups A and B under or without treatment.

of Williams et al. (15). For early childhood (18 –36 months),

P ⬍ 0. 001, comparisons between infants of groups A and B under or without treatment.

P ⬍ 0.05, comparisons between infants of groups A and B under or without treatment.
reference ranges for T4, FT4, T3, and TSH are 6 –14.8 ␮g/dl
(77.3–190.6 nmol/liter), 8.5–17.5 pg/ml (10.9 –22.5 pmol/liter),
83–250 ng/dl (1.2–3.75 nmol/liter), and 0.3–5 mU/liter, respec-
tively. Reference ranges were established in our own laboratory.
2591 ⫾ 809.9 (53)b

178.1 ⫾ 63.54 (46)b

Under treatment

46.65 ⫾ 34.7 (54)a

27.73 ⫾ 29.5 (54)a

8.49 ⫾ 3.2 (52)a
10.63 ⫾ 3.8 (36)a

Statistical analysis
25 (47.17)a
(n ⴝ 54)

The SPSS statistical software (version 17.0; Chicago, IL) was

230 (40)d

whole blood ⫻ 2.2 ⫽ milliunits per liter serum. ns, Not significant.

used, and the statistical significance was set at P ⫽ 0.05. The

Student t test was used for comparison of numerical variables.
Pearson ␹2 was used for the comparison of proportions.

Ethical aspects
Approval for this study was obtained from the Institute of
TSH (mU/liter serum), mean ⫾ SD (n) (P)

Child Health’s local ethics committee, and parents or children’s

Urine iodine (␮g/liter), median (n) (P)

TABLE 1. Infants with CH and the recall rate in

FT4 (pg/ml), mean ⫾ SD (n) (P)

relation to various cutoff points of TSH in 311,390

T4 (␮g/dl), mean ⫾ SD (n) (P)

T3 (ng/dl), mean ⫾ SD (n) (P)

screened newborns during a 3-yr period (January 2000

Age (d), mean ⫾ SD (n) (P)
Parameters

BW (g), mean ⫾ SD (n) (P)

BW ⬍2500 g (n) (%) (P)

to December 2002)
TSH cutoff Newborns Recall Infants
(mU/liter whole blood) recalled, n rate (%) with CH, n
30 173 0.05 114
TABLE 2.

20 376 0.12 144

P value ns.

10 3784 1.20 200

Conversion factor: neonatal TSH in milliunits per liter whole blood ⫻
2.2 ⫽ milliunits per liter serum.
b

d
a

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 07 June 2015. at 13:34 For personal use only. No other uses without permission. . All rights reserved.
4286 Mengreli et al. Screening for Congenital Hypothyroidism J Clin Endocrinol Metab, September 2010, 95(9):4283– 4290

legal escorts gave informed consent for the child to participate in concentration between 10 and 20 mU/liter whole blood
the study. and group BR included those with neonatal TSH value
greater than 20 mU/liter whole blood. Almost one third
(29.8%) of the infants with permanent CH (40 of 134)
Results belonged to group AR.
There was no significant difference in presence of pos-
In Table 1 the number of recalled newborns, recall rate,
itive thyroid antibodies between children with transient
and number of infants with CH using various TSH cutoff
points are presented. The number of infants given re- and permanent CH.
placement therapy increased by 28% (56 of 200) as the The hormonal profile of all CH infants at diagnosis
cutoff point decreased from 20 to 10 mU/liter whole (groups A and B) and reevaluation (groups AR and BR) is
blood. At the same time, a 10-fold increase in recall rate shown in Fig. 1.
was observed. Thirty-one of the reevaluated infants with CH were
In Table 2 comparisons of various parameters of in- premature (20.5%). Four had a gestational age less than
fants who had received treatment for CH and those whose 30 wk, 15 between 31 and 33 wk, and the rest had a
TSH had been normalized by the time of confirmation of gestational age between 34 and 36 wk. In Table 3 it is
diagnosis (false positives) are presented, according to the shown that 23 of 31 premature babies suffered permanent
screening cutoff point of TSH: group A, infants with neo- CH and more than half of them (15 of 23) belonged to
natal TSH values between 10 and 20 mU/liter whole blood group AR (65.2%), a percentage significantly higher com-
and group B, infants with neonatal TSH values greater pared with that of full-term infants (P ⫽ 0.026).
than 20 mU/liter whole blood. In group A all but one At reevaluation serum TSH concentration did not sig-
(serum TSH) parameters between infants treated and un- nificantly differ between preterm and full-term infants ei-
treated were comparable. On the contrary, in group B age ther in group AR or BR, although TSH levels were shifted
at diagnosis, serum T4, FT4, and T3 were significantly toward lower values in premature babies (Figs. 1, TSH
lower in the infants treated compared with those not graph, and 2).
treated. The percentage of recalled infants (either with or In Table 4 the etiologic classification of CH is pre-
without CH) with birth weight (BW) less than 2500 g was sented. In group AR structural abnormalities of the thy-
significantly higher in group A compared with that of roid gland (thyroid dysgenesis in situ or ectopic thyroid
group B. gland) were detected in 20% of CH cases (eight of 40) and
Thyroid antibodies were measured in 164 of 200 CH in group BR in 66% (62 of 94). Functional defects of the
children and 11 were found positive in either one or both thyroid were the predominant cause of CH in group AR
of them (6.7%). (80%), whereas it was 34% in group BR. Partial iodide
One hundred fifty-one of 200 CH infants (75.5%) were organification defects were found in two of 32 children in
reevaluated as previously described. The remaining in- the former group (20 and 38% discharge, respectively)
fants were not included in the study because they were and four of 32 in the latter group (16, 25, 76, and 78%
followed by other pediatric endocrine units closer to their discharge, respectively). In the latter group, one case with
native home address. total iodide organification defect was also detected. Four-
One hundred thirty-four of the 151 reevaluated chil- teen of 32 cases (43.7%) were familial in group AR and
dren (88.7%) suffered permanent CH (Table 3). seven of 32 in group BR (21.9%).
All reevaluated infants were divided into two groups: In the present study, by applying a TSH threshold of 10
group AR included the CH infants with neonatal TSH mU/liter whole blood, the estimated overall prevalence of

TABLE 3. Reevaluated infants (full-term and premature) with permanent and transient CH in relation to TSH cutoff
value at screening
Group AR (n ⴝ 47) Group BR (n ⴝ 104)
Infants with Infants with Infants with Infants with
permanent CH, n transient CH, n permanent CH, n transient CH, n
Reevaluated infants (n ⫽ 151) 40 7 94 10
Full-term infants (n ⫽ 120) 25 4 86 5
Premature infants (n ⫽ 31) 15 3 8 5
Group AR, Reevaluated infants with neonatal TSH levels between 10 and 20 mU/liter whole blood; group BR, reevaluated infants with neonatal
TSH levels greater than 20 mU/liter whole blood. Conversion factor: neonatal TSH milliunits per liter whole blood ⫻ 2.2 ⫽ milliunits per liter
serum. (For the reevaluation process, see section on methodology.)

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 07 June 2015. at 13:34 For personal use only. No other uses without permission. . All rights reserved.
J Clin Endocrinol Metab, September 2010, 95(9):4283– 4290 jcem.endojournals.org 4287

permanent CH is 1:1,749, whereas that of

transient CH 1:14,154 children.

Discussion
During the study period, using a TSH cutoff
point of 10 mU/liter whole blood, 56 addi-
tional infants with CH were diagnosed (16).
At follow-up the majority of those reevalu-
ated suffered permanent CH (88.7%), a
finding indicating that a mild increase in
screening TSH is not necessarily a predictor
of transient CH.
Analysis of the etiology of permanent CH
in the children with screening TSH values
between 10 and 20 mU/liter (group AR) and
those with TSH greater than 20 mU/liter
(group BR) shows that functional defects of
thyroid gland constitute the main cause of
thyroid dysfunction in the former (80%),
whereas it is less frequent (34%) in the latter
group. This finding is in agreement with the
results of a recent study by Corbetta et al.
(17), who reported an overall increase in
percentage of inborn functional defects in
CH children from 44 to 68% by applying
lower cutoff points in their screening pro-
gram in Lombardy. They assumed that this
milder expression of thyroid hypofunction
is due to genetic defects that cause perma-
nent nonautoimmune thyroid dysfunction
later in life (6, 18 –21).
The remaining 20% of children with per-
manent CH in group AR had anatomical
defects of the thyroid gland; four had thy-
roid hemiagenesis (10.0%), two had ectopic
thyroid tissues, and another two showed
asymmetry of the two lobes with increased
uptake by the right one. The first two dis-
orders, if untreated, lead to compensatory
hypertrophy of the thyroid tissue present in
the majority of such cases (22, 23).
The hormonal profile of children with CH
FIG. 1. The hormonal profile of children with permanent CH at diagnosis (groups A and of group A shows that a higher number of
B) and at reevaluation (groups AR and BR). Groups A and AR, CH infants with neonatal them had thyroid hormone values in the hy-
TSH levels between 10 and 20 mU/liter whole blood; groups B and BR, CH infants with pothyroid region by the time the diagnosis of
neonatal TSH levels greater than 20 mU/liter whole blood. (Subgroups AR1 and BR1 are
premature and AR2 and BR2 full-term infants) (conversion factor: neonatal TSH in CH was made than during reevaluation pe-
milliunits per liter whole blood ⫻ 2.2 ⫽ milliunits per liter serum). (For the reevaluation riod, a finding that may be attributed to the
process, see section on methodology). increased thyroid hormone requirements
for growth and development in early life.
It is known that prematurity complicates screening for
primary hypothyroidism due to developmental delay in

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 07 June 2015. at 13:34 For personal use only. No other uses without permission. . All rights reserved.
4288 Mengreli et al. Screening for Congenital Hypothyroidism J Clin Endocrinol Metab, September 2010, 95(9):4283– 4290

compared with full-term infants (25.8 and 7.5%, respec-

tively). Interestingly, the percentage of preterm infants
with transient CH was higher in group BR compared with
group AR (38.5 vs. 16.7%), indicating an inability of their
thyroid to cope with external stimuli, such as iodine over-
load due to antiseptics, drugs, etc.
In the past, when CH was diagnosed on clinical grounds,
its incidence was estimated to be around 1:6000 –7000 chil-
dren (28). The operation of screening programs revealed a
higher incidence of CH (1:3000 – 4000 children). Recent
publications refer to an even higher incidence for different
known or unknown reasons. This increase may be ex-
FIG. 2. The absolute serum TSH values at reevaluation in groups AR
and BR. Group AR, CH infants with neonatal TSH levels between 10 plained by improvement in the specificity of screening pro-
and 20 mU/liter whole blood; group BR, CH infants with neonatal TSH grams for CH, which use more than one analyte (T4-TSH-
levels greater than 20 mU/liter whole blood. (Subgroups AR1 and BR1 T4-binding globulin) (29), by higher CH incidence in twins
are premature and AR2 and BR2 full-term infants) (conversion factor:
neonatal TSH milliunits per liter whole blood ⫻ 2.2 ⫽ milliunits per
or triplets because the number of medically induced preg-
liter serum). nancies is increasing (30, 31), or by other unknown factors
(32). Another, rather significant cause for the higher num-
the maturation of the hypothalamic-pituitary-thyroidal ber of children with CH depicted through neonatal screen-
axis, and different methodological approaches have been ing programs is the lower threshold level applied (17, 33,
proposed so that no preterm newborn escapes early diag- 34). In this study the prevalence of permanent CH is higher
nosis for CH, i.e. repeat specimen collection (24 –26). In than that previously reported by us (1:1749 vs. 1:3384) by
the present study, the lowering of TSH cutoff point from using the same reagents but having a lower TSH threshold.
20 to 10 mU/liter resulted in a significant increase in the It is known that changes in threshold limits influence
number of recalled newborns with BW less than 2500 g the number of false-positive and false-negative results
(37.5%). The diagnosis of permanent CH was made in 15 (34). The lowering of the cutoff point by 10 mU/liter has
of the 18 reevaluated premature (83.3%). Korada et al. increased 10 times the number of children requiring re-
(27), similarly, reported that they picked up a preterm evaluation in whom the diagnosis of hypothyroidism was
baby with CH by lowering the neonatal TSH threshold not confirmed (false positive results). This is a drawback
from 10 to 6 mU/liter (TSH was measured by a fluoroim- of screening with adverse economic and emotional effects
munometric assay), and they suggested that a second TSH to the parents. The additional cost from the increased
screen in preterm infants may be unnecessary if a lower TSH number of recalled infants amounts to about 1.8% of the
threshold is used. screening budget.
With regard to transient CH, the overall percentage As in other endocrinometabolic disorders, the hor-
was 11.2% and it was significantly higher in premature monal and clinical phenotype of CH is a continuous spec-

TABLE 4. Etiologic classification of the children with permanent CH

Scintiscan results Group AR, n (%) Group BR, n (%)
Functional defects 32 (80%) 32 (34.0%)
Normal thyroid gland 27 (13 of 27 familial) 19 (4 of 19 familial)
Thyroid gland with
Positive perchlorate discharge test 2 5 (3 of 5 familial)
Low iodine uptake 1 (familial) 2
Increased size 2 6
Anatomical defects (developmental disorders) 8 (20%) 62 (66.0%)
Athyreosis 0 24
Thyroid dysgenesis in situ
Left lobe absent or hypoplastic 4 1
Asymmetry of the two lobes with increased uptake by the right one 2 0
Hypoplastic thyroid gland 0 1
Ectopic thyroid tissue (lingual thyroid) 2 36
Total number of infants 40 94
Group AR, Reevaluated infants with neonatal TSH between 10 and 20 mU/liter whole blood; group BR, reevaluated infants with neonatal TSH
levels greater than 20 mU/liter whole blood. Conversion factor: neonatal TSH milliunits per liter whole blood ⫻ 2.2 ⫽ milliunits per liter serum. (For
the reevaluation process see section on methodology.)

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 07 June 2015. at 13:34 For personal use only. No other uses without permission. . All rights reserved.
J Clin Endocrinol Metab, September 2010, 95(9):4283– 4290 jcem.endojournals.org 4289

trum expanding from severe hypofunction of thyroid 7. Gillis D, Brnjac L, Perlman K, Sochett EB, Daneman D 1998 Fre-
quency and characteristics of lingual thyroid not detected by screen-
gland expressed with high levels of TSH and very low
ing. J Pediatr Endocrinol Metab 11:229 –233
concentration of thyroid hormones to subclinical or com- 8. Mengreli C, Kassiou K, Tsagaraki S, Pantelakis S 1981 Neonatal
pensated thyroid dysfunction, in which mildly elevated screening for hypothyroidism in Greece. Eur J Pediatr 137:185–187
TSH levels and normal thyroid hormones concentrations 9. 1999 Revised guidelines for neonatal screening programmes for pri-
mary congenital hypothyroidism. Working Group on Neonatal
are found. One may pose the following question: should a Screening of the European Society for Paediatric Endocrinology.
screening program be considered satisfactory if it discov- Horm Res 52:49 –52
ers the severe forms of CH that definitely lead to mental 10. Panoutsopoulos G, Mengreli C, Ilias I, Batsakis C, Christakopoulou
I 2001 Scintigraphic evaluation of primary congenital hypothyroid-
retardation, or should it also aim at identifying subtle
ism: results of the Greek screening program. Eur J Nucl Med 28:
changes of thyroid function, which will potentially affect 529 –533
cognition, growth, and abnormalities of lipid metabolism 11. Delange F 1997 Neonatal screening for congenital hypothyroidism.
and cardiac function? Results and perspectives. Horm Res 48:51–56
12. LaFranchi S 1999 Congenital hypothyroidism: etiologies, diagnosis,
Obviously there are unanswered questions concerning and management. Thyroid 9:735–740
the benefit with regard to the long-term outcome of chil- 13. Pino S, Fang SL, Braverman LE 1996 Ammonium persulfate: a safe
dren with mild hypothyroidism treated early (35), and alternative oxidizing reagent for measuring urinary iodine. Clin
Chem 42:239 –243
only well designed prospective studies could clarify this 14. World Health Organization, United Nations Children’s Fund,
important issue. International Council for the Control of Iodine Deficiency Disorders
In conclusion, we demonstrated that a considerable (WHO/UNICEF/ICCIDD) 2007 Assessment of the iodine defi-
ciency disorders and monitoring their elimination. A guide for pro-
number of infants, especially premature, with CH escape
gramme managers. 3rd ed. Geneva: World Health Organization
diagnosis on a neonatal screening applying a TSH cutoff 15. Williams FL, Simpson J, Delahunty C, Ogston SA, Bongers-Schokking
point of 20 mU/liter whole blood. Being aware of the re- JJ, Murphy N, van Toor H, Wu SY, Visser TJ, Hume R; Collaboration
sults of this prospective study, we have considered a mod- from the Scottish Preterm Thyroid Group 2004 Developmental trends
in cord and postpartum serum thyroid hormones in preterm infants.
ification of our screening methodology and follow-up. J Clin Endocrinol Metab 89:5314 –5320
16. Mengreli C, Magiakou AM, Girginoudis P, Kanaka-Gantenbein C,
Dacou-Voutetakis C 2007 Congenital hypothyroidism (CH): a sig-
nificant number of false negative results when a cutoff point of 20
Acknowledgments mU/L is applied. Horm Res 68(Suppl 1):218
17. Corbetta C, Weber G, Cortinovis F, Calebiro D, Passoni A, Vigone
We thank Kiki Koutsoukou, Niki Vanikioti, Stella Manthou, MC, Beck-Peccoz P, Chiumello G, Persani L 2009 A 7-year expe-
and Sophia Dakoutrou for technical assistance. We also appre- rience with low blood TSH cutoff levels for neonatal screening re-
ciate Ioannis Monopolis for assistance with statistical analysis veals an unsuspected frequency of congenital hypothyroidism. Clin
and data management. Endocrinol (Oxf) 71:739 –745
18. Alberti L, Proverbio MC, Costagliola S, Romoli R, Boldrighini B,
Vigone MC, Weber G, Chiumello G, Beck-Peccoz P, Persani L 2002
Address all correspondence and requests for reprints to: Germline mutations of TSH receptor gene as cause of nonautoim-
Chryssanthi Mengreli, Ph.D., Department of Biochemical Lab- mune subclinical hypothyroidism. J Clin Endocrinol Metab 87:
oratories, Institute of Child Health, Aghia Sophia Children’s 2549 –2555
Hospital, 11527-Goudi, Athens, Greece. E-mail: chmengreli@ 19. Moreno JC, de Vijlder JJ, Vulsma T, Ris-Stalpers C 2003 Genetic
ich.gr. basis of hypothyroidism: recent advances, gaps and strategies for
Disclosure Summary: The authors have nothing to disclose. future research. Trends Endocrinol Metab 14:318 –326
20. Park SM, Chatterjee VK 2005 Genetics of congenital hypothyroid-
ism. J Med Genet 42:379 –389
21. Zamproni I, Grasberger H, Cortinovis F, Vigone MC, Chiumello G,
References Mora S, Onigata K, Fugazzola L, Refetoff S, Persani L, Weber G
2008 Biallelic inactivation of the dual oxidase maturation factor 2
1. Burrow GN, Dussault JH, eds. 1980 Neonatal thyroid screening. (DUOXA2) gene as a novel cause of congenital hypothyroidism.
New York: Raven Press J Clin Endocrinol Metab 93:605– 610
2. Leutwyler K 1995 The price of prevention. Sci Am 272:124 –129 22. Maiorana R, Carta A, Floriddia G, Leonardi D, Buscema M, Sava
3. Loeber JG 2007 Neonatal screening in Europe: the situation in 2004. L, Calaciura F, Vigneri R 2003 Thyroid hemiagenesis: prevalence in
J Inherit Metab Dis 30:430 – 438 normal children and effect on thyroid function: J Clin Endocrinol
4. de Zegher F, Vanderschueren-Lodeweyckx M, Heinrichs C, Van Metab 88:1534 –1536
Vliet G, Malvaux P 1992 Thyroid dyshormonogenesis: severe hy- 23. Kuehnen P, Grueters A, Krude H 2009 Two puzzling cases of thyroid
pothyroidism after normal neonatal thyroid stimulating hormone dysgenesis. Horm Res 71(Suppl 1):93–97
screening. Acta Paediatr 81:274 –276 24. Mitchell ML, Walraven C, Rojas DA, McIntosh KF, Hermos RJ
5. Vincent MA, Rodd C, Dussault JH, Van Vliet G 2002 Very low birth 1994 Screening very low birth weight infants for congenital hypo-
weight newborns do no need repeat screening for congenital hypo- thyroidism. Lancet 343:60 – 61
thyroidism. J Pediatr 140:311–314 25. Hunter MK, Mandel SH, Sesser DE, Miyabira RS, Rien L, Skeels
6. Moreno JC, Klootwijk W, van Toor H, Pinto G, D’Alessandro M, MR, LaFranchi SH 1998 Follow-up of newborns with low thyroxine
Lèger A, Goudie D, Polak M, Grüters A, Visser TJ 2008 Mutations and nonelevated thyroid-stimulating hormone-screening concentra-
in the iodotyrosine deiodinase gene and hypothyroidism. N Engl tions: results of the 20-year experience in the Northwest Regional
J Med 358:1811–1818 Newborn Screening Program. J Pediatr 132:70 –74

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 07 June 2015. at 13:34 For personal use only. No other uses without permission. . All rights reserved.
4290 Mengreli et al. Screening for Congenital Hypothyroidism J Clin Endocrinol Metab, September 2010, 95(9):4283– 4290

26. Mandel SJ, Hermos RJ, Larson CA, Prigozhin AB, Rojas DA, Mitchell roidism in multiple pregnancies. J Clin Endocrinol Metab 92:3141–
ML 2000 Atypical hypothyroidism and the very low birth weight in- 3147
fant. Thyroid 10:693– 695 31. Sakka SD, Malamitsi-Puchner A, Loutradis D, Chrousos GP,
27. Korada M, Pearce MS, Ward Platt MP, Avis E, Turner S, Wastell H, Kanaka-Gantenbein C 2009 Euthyroid hyperthyrotropinemia in
Cheetham T 2008 Repeat testing for congenital hypothyroidism in children born after in vitro fertilization. J Clin Endocrinol Metab
preterm infants is unnecessary with an appropriate thyroid stimu- 94:1338 –1341
lating hormone threshold. Arch Dis Child Fetal Neonatal Ed 93: 32. Harris KB, Pass KA 2007 Increase in congenital hypothyroidism in
F286 –F288 New York State and in the United States. Mol Genet Metab 91:
28. Alm J, Hagenfeldt L, Larsson A, Lundberg K 1984 Incidence of 268 –277
congenital hypothyroidism: retrospective study of neonatal labora- 33. Pryce RA, Gregory JW, Warner JT, John R, Bradley D, Evans C
tory screening versus clinical symptoms as indicators leading to di- 2007 Is the current threshold level for screening for congenital hy-
agnosis. Br Med J 289:1171–1175 pothyroidism too high? An audit of the clinical evaluation, confir-
29. Kempers MJ, Lanting CI, van Heijst AF, van Trotsenburg AS, matory diagnostic tests and treatment of infants with increased
Wiedijk BM, de Vijlder JJ, Vulsma T 2006 Neonatal screening for blood spot thyroid-stimulating hormone concentrations identified
congenital hypothyroidism based on thyroxine, thyrotropin, and on newborn blood spot screening in Wales. Arch Dis child 92:1048
thyroxine-binding globulin measurement: potentials and pitfalls. 34. Korada SM, Pearce M, Ward Platt MP, Avis E, Turner S, Wastell H,
J Clin Endocrinol Metab 91:3370 –3376 Cheetham T 2010 Difficulties in selecting an appropriate neonatal
30. Olivieri A, Medda E, De Angelis S, Valensise H, De Felice M, Fazzini TSH screening threshold. Arch Dis Child 95:169 –173
C, Cascino I, Cordeddu V, Sorcini M, Stazi MA; Study Group for 35. Rapaport R 2000 Congenital hypothyroidism: expanding the spec-
Congenital Hypothyroidism 2007 High risk of congenital hypothy- trum. J Pediatr 136:10 –12

Members receive
an email digest of endocrinology-related news
selected from thousands of sources.

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 07 June 2015. at 13:34 For personal use only. No other uses without permission. . All rights reserved.