http://informahealthcare.com/autISSN: 0891-6934 (print), 1607-842X (electronic)

Autoimmunity, Early Online: 1–7! 2014 Informa UK Ltd. DOI: 10.3109/08916934.2014.914175

ORIGINAL ARTICLE

The clinical role of serum concentrations of selected cytokines: IL-1b,TNF-a and IL-6 in diagnosis of autoimmune thyroid disease (AITD) inchildren

Hanna Mikos1,2, Marcin Mikos3, Barbara Rabska-Pietrzak1, and Marek Niedziela1,2

1Department of Pediatric Endocrinology and Rheumatology, 2Molecular Endocrinology Laboratory, and 3Department of Pneumonology,

Allergology and Clinical Immunology, Poznan University of Medical Sciences, Poznan, Poland

Abstract

Chronic autoimmune thyroiditis (cAIT) leads to hypothyroidism due to T cell-mediatedcytotoxicity in most cases. By contrast, Graves’ disease (GD) with thyrotropin receptorstimulatory autoantibodies cause hyperthyroidism. Cytokines play a crucial role in modulatingimmune response in both disorders. The aim of study was to evaluate the concentrations ofcytokines: IL-1b, TNF-a and IL-6 in these two opposite clinical and hormonal thyroid diseases.The study group consisted of 64 children, 44 newly diagnosed, untreated children with cAIT(n¼ 22; with hypothyroidism) and GD (n¼ 22; hyperthyroidism), and the control group of 20healthy children. Cytokine concentrations were evaluated using the ELISA technique. Thestudied groups of children did not differ significantly in concentrations of IL-6 (p¼ 0.48) andTNF-a (p¼ 0.067). In children with hypothyroidism, we found significantly higher concentra-tions of IL-1b (median 2.16 pg/ml, IQR 0.87) compared to hyperthyroidism (median 1.39 pg/l,IQR 1.27) (p50.01) and the control group (median 1.88 pg/ml, IQR 1.04) (p50.05). The results ofROC curve analysis demonstrated the usefulness of IL-1b (AUC¼ 0.77, p¼ 0.003) and TNF-a(AUC¼ 0.691, p¼ 0.034) as diagnostic parameters in cAIT which enable discrimination ofchildren with autoimmune thyroid disease from healthy individuals. Concentrations of thesemarkers are increased in autoimmune hypothyroidism. We found no significant sex differencesin the tested parameters. In conclusion, IL-1b and TNF-a may be considered as markers ofhypothyroidism, and could efficiently discriminate between healthy and autoimmunehypothyroid children. Significantly higher concentrations of IL-1b in children with hypothy-roidism may be used to distinguish children with cAIT from GD patients.

Keywords

Autoimmunity, cytokines, children, serum,thyroid

History

Received 15 February 2014Revised 22 March 2014Accepted 8 April 2014Published online 7 May 2014

Introduction

Autoimmune thyroid disease (AITD) is the most common

organ-specific autoimmune disease. It is characterized by

infiltration of the gland by T- and B-lymphocytes, dendritic

cells and macrophages. The presence of autoantibodies

reacting against thyroid autoantigens and alterations in T

cell subsets are also characteristic features of AITD [1].

Autoimmune thyroid dysfunction creates an image consisting

of two opposing clinical and hormonal states of thyroid

function: chronic autoimmune thyroiditis (cAIT) presenting

as hypothyroidism and Graves’ disease (GD), characterized by

hyperthyroidism. The thyroid gland of patients with auto-

immune thyroid disease is morphologically characterized by

massive infiltration of lymphoid cells. The interleukin-1

(IL-1), tumor necrosis factor-a (TNF-a) and interleukin-6

(IL-6) are involved in the intercellular communication

required to initiate and control immune response. However,

their role in the pathogenesis of AITD is unclear. Antigen

determinants activate naive CD4+ T cells, which may

differentiate into Th1 cells and contribute to the development

of HT [2]. Reduced number of Tregs with CD4+CD25 and

CD4+FoxP3 phenotype suggests their role in initiation of

autoimmune process in thyroid disorders in children [3]. Both

T helper 17 (Th17) and T regulatory cells might play a role in

the pathogenesis of GD [4]. B cells have taken the center

stage in tolerance and autoimmunity, mainly through polar-

ization of T cell subsets (Tregs/Th1/Th2) and their ability to

produce cytokines [5]. The pathogenesis of GD is complex

and heterogeneous and hyperthyroidism is an example of

antibody-mediated organ-specific autoimmunity. Since TSAb

and B-cell proliferation is hallmark of GD, Th2 responses

have been associated with its pathogenesis [6]. Cytokines,

predominantly IL-1b, TNF-a and IL-6, have been implicated

in the pathogenesis of AITD and are also considered to play

important role in intrathyroidal inflammation and modulation

of thyrocyte growth and function [7,8].

Correspondence: Prof. Marek Niedziela, MD, PhD, Department ofPediatric Endocrinology and Rheumatology, 2nd Chair of Pediatrics,Poznan University of Medical Sciences, 27/33, Szpitalna Street, 60-572Poznan, Poland. Tel: +48 61 8491481. Fax: +48 61 8480291. E-mail:mniedzie@ump.edu.pl

Aut

oim

mun

ity D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y T

ham

mas

at U

nive

rsity

on

10/0

8/14

For

pers

onal

use

onl

y.

Materials and methods

Patients

The study group consisted of 64 children: 45 girls (70.3%) and

19 boys (23.7%). The study group was divided into three

subgroups: 22 newly diagnosed, untreated hypothyroid chil-

dren with cAIT, including 18 girls (81.8%) and 4 boys

(18.2%), with median age of 11.3 years (range: 6.7–15.9); 22

untreated hyperthyroid children with GD: including 16 girls

(72.7%) and 6 boys (27.3%), with median age of 14.7 years,

(range: 3.2–18.1) and a control group of 20 healthy children:

11 girls (55%) and 9 boys (45%), with median age of 12.7

years (range: 1.9–18.0). These groups were comparable in

terms of age and gender distribution (p¼ 0.31). Therefore, the

study included children with newly diagnosed autoimmune

thyroid disease. Patients were recruited from the Department

of Endocrinology and Rheumatology K. Jonscher’s Clinical

Hospital, Poznan University of Medical Sciences and pedi-

atric endocrinology outpatient clinic. All children underwent

full history, pediatric examination, laboratory tests and

thyroid ultrasound. The study was performed in accordance

with the Helsinki Declaration and good clinical practice and

was approved by the local Ethics Committee. The inclusion

criteria were: female and male subjects, age below 18 years of

age and confirmed diagnosis cAIT and GD (criteria: hormo-

nal, immunological profile and thyroid ultrasound).

Thyroid function

Serum samples were tested for serum concentrations of TSH

(reference range: 0.46–4.67 mIU/ml) and free T4 (reference

range: 0.71–1.85 ng/dl), free T3 (reference range: 1.45–

3.48 pg/ml) using Abbott AxSYM System (Abbott

Diagnostics, Germany). Antithyroid antibodies (ATG,

ATPO, TRAb) in all participants were examined by radio-

immunoassay, according to the manufacturers’ instructions

(BRAHMS, Berlin, Germany). The reference values for

ATPO and ATG were below 60 IU/ml and for TRAb below

1 U/l. The tests were conducted in the Laboratory of the K.

Jonscher’s Clinical Hospital of Poznan University of Medical

Sciences. Anthropometrical measurements (weight and

height) were performed in all participants of the study and

were used to calculate BMI, BMI-SDS and Cole index.

Reference percentile charts for Polish children were used for

the analysis of the obtained data [9] (Table 1).

Cytokine assays

Peripheral blood samples were collected before treatment

introduction. Sera were collected after centrifugation imme-

diately after blood was drawn and stored at �70 �C until

assayed. Serum concentrations of IL-1b, TNF-a and IL-6

were evaluated using ELISA kits (Bender MedSystems

GmbH, Vienna, Austria) according to manufacturer’s instruc-

tions. The limit of detection for IL-1b was 0.32 pg/ml (intra-

and interassay CV: 8.6% and 5.1%, respectively); for TNF-a:

5.0 pg/ml (intra- and interassay CV: 8.1%, 7.7%); for IL-6:

0.92 pg/ml (intra- and interassay CV: 5.2%, 3.2%). The

analyses were conducted in the Molecular Endocrinology

Laboratory of the Department of Pediatric Endocrinology and

Rheumatology of Poznan University of Medical Sciences

(Table 2).

Statistical analyses

A Shapiro–Wilk test was used to verify the normality of

subsets of data. Descriptive statistics including mean with

standard deviation for normally distributed variables was

reported, otherwise median with interquartile range was

reported. The differences between the two groups were

analyzed for significance either with t-test or Mann–Whitney

non-parametric test. The significance of differences between

all three groups was either tested with ANOVA with

Newman–Keuls post-test or with a Kruskal–Wallis test

with Dunn’s post-test. Nominal variables were tested with

Pearson’s chi-square. The correlations between variables were

evaluated by Spearman rank correlation test. To determine the

diagnostic value of cytokines, we calculated the area under

the curve (AUC) of the receiver operating characteristic

(ROC) curve. Cut-off values for ROC curves were calculated

with J-Youden statistic. SPSS 17.0 for Windows (SPSS,

Chicago, IL) was used in the analysis. A p value less than 0.05

was considered statistically significant.

Results

Relationships between metabolic and thyroidfunction

We found no significant differences in the tested parameters

between boys and girls in groups of children with cAIT and

GD (age, BMI, BMI-SDS, Cole index, TSH, fT4, fT3, ATG,

ATPO, TRAb, IL-1b, TNF-a and IL-6). There were no

Table 1. Demographic and clinical characteristics of groups.

HypothyroidismMean ± SD median (IQR)

HyperthyroidismMean ± SD median (IQR)

Control groupMean ± SD median (IQR) s

n n¼ 22 n¼ 22 n¼ 20

Sex 18 girls/4 boys 16 girls/6 boys 11 girls/9 boys nsAge (years) 11.39 ± 2.68 14.75 (7.09) 11.57 ± 4.86 nsBMI (kg/m2) 18.69 (5.45) 18.25 ± 3.42 18.17 ± 3.50 nsBMI-SDS 0.3 (2.04) �0.38 ± 1.05 �0.55 (1.29) nsCole index 1.05 ± 0.22 0.95 ± 0.13 0.9 ± 70.14 nsTSH (0.5–5.0mIU/ml) 37.34 (17.69)" 0 (0.01)# 2.42 (1.52) p50.001 (K–W)fT4 (0.7–1.85 ng/dl) 0.54 ± 0.31# 4.24 ± 1.06" 1.03 ± 0.12 p50.001 (ANOVA)fT3 (1.7–3.5 pg/ml) 2.10 ± 0.97 19.01 ± 5.30" 2.70 ± 0.56 p50.001 (ANOVA)ATG (560 IU/ml) 124 (589)" 101 (552)" 20 (4.5) p50.001 (K–W)ATPO (560 IU/ml) 3000 (111)" 3000 (1536)" 10 (29) p50.001 (K–W)TRAb (51 IU/l) 0.7 ± 0.3 16.75 (24.6)" 0.5 ± 0.3 p50.001 (K–W)

#: below reference range; ": above reference range.

2 H. Mikos et al. Autoimmunity, Early Online: 1–7

Aut

oim

mun

ity D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y T

ham

mas

at U

nive

rsity

on

10/0

8/14

For

pers

onal

use

onl

y.

significant differences in age and nutritional status param-

eters: BMI, BMI-SDS and Cole index between the groups

with cAIT, GD and control (p¼ 0.31, p¼ 0.8, p¼ 0.31,

p¼ 0.15). The study identified statistically significant differ-

ences (p50.001) of parameters of thyroid hormonal function

(TSH, fT4 and fT3) and immune markers: (ATG, ATPO and

TRAb) between the analyzed groups of children. All of these

observed differences are directly related to the opposing states

of hormonal hyperthyroidism and hypothyroidism compared

to the control group.

Cytokine serum concentrations

Serum IL-6 levels did not differ significantly between all

groups: cAIT (median; IQR; range: 1.60 pg/ml; 0.07; 0.39–

11.10), GD (1.60 pg/ml; 0.04; 0.39–9.18) and control

(1.56 pg/ml; 0.37; 0.12–2.48) (p¼ 0.48). Likewise, no sig-

nificant differences in TNF-a concentration between the

groups of children were found: cAIT (median; IQR; range:

15.08 pg/ml; 21.94; 0.00–143.44), GD (13.63 pg/ml; 15.28;

0.00–28.51) and control (0.96 pg/ml; 12.81; 0.00–17.75)

(p¼ 0.067).

Significantly higher serum IL-1b concentration was found

in children with cAIT (median; IQR; range: 2.16 pg/ml; 0.87;

0.11–9.53), compared to control (1.88 pg/ml; 1.04; 0.01–2.38)

(p50.05) and cAIT compared to GD (1.39 pg/ml; 1.27; 0.26–

2.68) (p50.01) (Figure 1). There was no difference in IL-1bconcentration between children with hyperthyroidism and

control group (p40.05).

Correlations of cytokine concentrations

Significant positive correlations between cytokines were

identified; in cAIT: IL-1b and TNF-a (r¼ 0.45, p50.05)

and in GD: IL-6 and TNF-a (r¼ 0.62, p50.01). TNF-apositively correlated with ATPO in cAIT (r¼ 0.54, p50.01)

as well as IL-1b and ATPO in GD (r¼ 0.47; p50.05). In the

control group, we found negative correlation of IL-1b and

TNF-a (r¼�0.58, p50.01).

ROC analysis

The results of ROC curve analysis enabled determination of

usefulness of monitoring cytokine concentrations in order to

discriminate children with autoimmune thyroid disease from

healthy individuals. Cytokines IL-1b (AUC¼ 0.77,

p¼ 0.003) (Table 3 and Figure 2) and TNF-a(AUC¼ 0.691, p¼ 0.034) (Table 3 and Figure 3) discrimi-

nated healthy and autoimmune-hypothyroid (cAIT) children

Table 2. Comparison of serum cytokines concentrations between groups.

HypothyroidismMean ± SD median (IQR)

HyperthyroidismMean ± SD median (IQR)

Control groupMean ± SD median (IQR)

n n¼ 22 n¼ 22 n¼ 20 s

TNF-a (pg/ml) 19.57 ± 31.95 10.44 ± 8.31 6.43 ± 7.26 ns15.80 (21.94) 13.63 (15.28) 0.96 (12.81) p¼ 0.067 (K–W)

IL-6 (pg/ml) 2.09 ± 2.13 1.96 ± 1.76 1.40 ± 0.53 ns2.60 (0.07) 1.60 (0.04) 1.56 (0.37) p¼ 0.48 (K–W)

IL-1b (pg/ml) 2.58 ± 1.80 1.45 ± 0.67 1.48 ± 0.702.16 (0.87) 1.39 (1.27) 1.88 (1.04) p¼ 0.002 (K–W)

Figure 1. Boxplot of IL-1b serum concentra-tions in the hypothyroid, hyperthyroid andcontrol groups. [*]: outlier.

DOI: 10.3109/08916934.2014.914175 IL-1�, TNF-� and IL-6 in AITD in children 3

Aut

oim

mun

ity D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y T

ham

mas

at U

nive

rsity

on

10/0

8/14

For

pers

onal

use

onl

y.

with low sensitivity of 59.1% and 54.5%, respectively, but

high specificity of 95% and 85% at cut-off points of 2.1 pg/ml

for IL-1b and 14.7 pg/ml for TNF-a. The concentration of

these markers increased in hypothyroidism. Moreover, the

ROC curve for IL-1b showed good sensitivity and specificity

for discriminating between the cAIT and GD groups:

AUC¼ 0.773; p¼ 0.002 (sensitivity 72.7%, specificity

86.4%) at cut-off point of 1.7 pg/ml (Table 4 and Figure 4).

None of the studied cytokines IL-1b, TNF-a and IL-6

effectively discriminated between GD and control (AUC did

not differ significantly from 0.5: p¼ 0.801, p¼ 0.092 and

p¼ 0.435, respectively) (Table 5).

Discussion

In AITD, the occurrence of both Th1 and Th2 cell-mediated

immune response is documented, but the specific role of

cytokines in both opposing clinical states of disease is not

clearly defined. cAIT is characterized by destruction of

thyroid cells and dominance of Th1 type immune response

with secretion of the following cytokines: IL-1b, IL-2, TNF-aand IFN-g. In contrast, GD is characterized by stimulation of

the gland through TSAb with Th2 cytokine pattern dominated

by secretion of IL-4, IL-5, IL-6, IL-9, IL-10 and IL-13. Th1

and Th2 cells are responsible for specific cytokine profiles in

Table 3. Area under the curve: cAIT versus control.

Asymptotic 95%confidence interval

Variable AreaStandard

errorAsymptoticsignificance

Lowerbound

Upperbound

IL-6 (pg/ml) 0.601 0.088 p¼ 0.262 (ns) 0.428 0.774IL-1b (pg/ml) 0.770 0.073 p¼ 0.003 0.627 0.914TNF-a (pg/ml) 0.691 0.082 p¼ 0.034 0.529 0.852

Figure 2. ROC curve of IL-1b: cAIT versus control group (AUC¼ 0.77,p¼ 0.003, cut-off¼ 2.1 pg/ml; sensitivity: 59.1%, specificity: 95%).

Table 4. Area under the curve: cAIT versus GD.

Asymptotic 95%confidence interval

Variable AreaStandard

errorAsymptoticsignificance

Lowerbound

Upperbound

IL-6 (pg/ml) 0.536 0.088 p¼ 0.681 (ns) 0.363 0.709TNF-a (pg/ml) 0.572 0.089 p¼ 0.411 (ns) 0.397 0.748IL-1b (pg/ml) 0.773 0.073 p¼ 0.002 0.629 0.916

Figure 3. ROC curve of TNF-a: cAIT versus control group(AUC¼ 0.691, p¼ 0.034, cut-off¼ 14.7 pg/ml, sensitivity: 54.5%, spe-cificity: 85%).

Figure 4. ROC curve of IL-1b: cAIT versus GD (AUC¼ 0.773,p¼ 0.002, cut-off¼ 1.7 pg/ml, sensitivity: 72.7%, specificity: 86.4%).

4 H. Mikos et al. Autoimmunity, Early Online: 1–7

Aut

oim

mun

ity D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y T

ham

mas

at U

nive

rsity

on

10/0

8/14

For

pers

onal

use

onl

y.

cAIT and GD. Th1 cells are involved in inflammatory

reactions and cellular immune responses. The corresponding

cytokines promote cytotoxic reactions and inflammatory

processes, participating in the delayed hypersensitivity reac-

tion in the tissue. Th2 cytokines stimulate production of

antibodies and are involved in the humoral immune response,

promoting immune and allergic reactions. The prevalence of

either Th1 or Th2 cytokine secretion varies from one disease

to another and leads to a different immune response.

Cytokines secreted by both T-cell populations exhibit an

inhibitory effect on each other and can shift the balance of

the immune response toward Th1 or Th2 cells [10].

In the present study we examined levels of of Th1

cytokines: IL-1b, TNF-a and Th2 cytokine: IL-6 in the serum

of children diagnosed with cAIT manifesting with hypothy-

roidism and in the serum of children with GD presenting

hyperthyroidism – prior to the introduction of treatment in

both diseases. The highest concentration of IL-1b was found

in children with cAIT (median of 2.16 pg/ml) and the lowest

in the GD group (median of 1.39 pg/ml), whereas in the

control group the median level of IL-1b was 1.88 pg/ml. The

concentration of IL-1b was significantly different between the

groups of children with cAIT and GD (p50.01) and in cAIT

versus control (p50.05). There was no difference in IL-1blevels between the group of children with GD and control

group (p40.05). The results confirm the important role of IL-

1b in the hypothyroid phase of chronic autoimmune

thyroiditis rather than in the severe phase of GD in children.

There were no significant differences of TNF-a and IL-6

levels between the studied groups of children (p¼ 0.067 and

p¼ 0.48, respectively).

Nanba and Watanabe [11] investigated the proportion of

peripheral Th1, Th2 and Th17 cells in patients with AITD.

The proportion of peripheral Th1 cells was higher in patients

with severe HT than in patients with mild HT (p50.05) and

the proportion of peripheral Th2 cells was lower in patients

with severe HT than in patients with mild HT (p50.001).

Therefore, the Th1/Th2 ratio was higher in severe HT in

comparison with mild HT (p50.001). The proportion of

peripheral Th17 cells in patients with AITD was higher than

in control subjects, and the proportion of these cells in

patients with intractable GD was higher than in patients with

GD in remission (p50.05). The authors concluded that the

peripheral Th1/Th2 cell ratio is related to the severity of HT,

and the proportion of Th17 cells is related to the intractability

of GD. They hypothesize that these patterns of peripheral Th

cell subsets may be expressed within the thyroid.

Rebuffat et al. [12] studied the effect of proinflammatory

cytokines IL-1b and TNF-a on the integrity of the epithe-

lium of thyrocytes in autoimmune thyroid diseases.

These cytokines stimulate secretion of IL-6 which may

modify epithelial integrity, transmembrane protein expression

and cell adhesion. High expression of these proteins in thyroid

follicular cells has been demonstrated in GD, in contrast to

the decreased expression of JAM-A and ZO-1 junction

proteins in cells derived from the thyroid gland of patients

suffering from HT. This is confirmed by the loss of the

integrity of the thyroid follicular cells, occurring as a result of

inflammatory processes in HT and apoptosis. It has been

shown that IL-1b modifies thyroid epithelial tightness –

altering the expression and localization of Claudin and ZO-1.

This confirms the important pathogenic role of IL-1b in

thyroid autoimmunity [1].

Our study confirmed a significant positive correlation of

IL-1b and TNF-a (r¼ 0.46, p50.05) in children with cAIT,

whereas in the control group the significant correlation of

IL-1b and TNF-a was negative (r¼ 0.58, p50.01). Both

elevated levels of IL-1b and a positive correlation profile

confirms the inflammatory nature of HT in children. ROC

analysis showed that IL-1b may aid in differentiation between

healthy children and patients with hypothyroidism in cAIT

(AUC¼ 0.77, p¼ 0.003) with low sensitivity (59.1%) but

very high specificity (95%) at a cut-off point of 2.1 pg/ml.

Thus, concentrations of IL-b below the cut-off point of 2.1 pg/

ml enable exclusion of cAIT in a given patient with high

probability. The ROC curve of cAIT versus GD for IL-1bshowed high sensitivity (72.7%) and high specificity (86.4%)

which demonstrates a similar ability to differentiate these

groups (AUC¼ 0.773, p¼ 0.002) at a lower cut-off point of

1.7 pg/ml compared to a ROC curve of cAIT versus control

(2.1 pg/ml). This result confirms that the serum level of IL-1bmay be an effective marker for hypothyroidism in chronic

autoimmune thyroiditis.

The presence of TNF-a in the serum of patients with

autoimmune thyroid disease may be an important stimulus for

the production of IL-6. Studies were conducted on the

cytokine profile of TNF-a and IL-6 in the serum of patients

with adult AITD in hypothyroidism and hyperthyroidism

before and after normalization of thyroid function [13]. The

data in the literature confirm the elevated levels of IL-6

and TNF-a observed in hyperthyroidism of Graves’ disease

[14–18] as well as in TAO [19].

Trials made by Pichler et al. [20] on 39 patients with newly

diagnosed Graves’ disease showed a higher IL-6 level only in

four cases compared to healthy controls.

Kiziltunc et al. [21] investigated a group of 27 adult

patients between 58 and 64 years of age with untreated

hyperthyroidism and 23 with untreated hypothyroidism. The

average concentrations of cytokines TNF-a and IL-6 were

comparable to a group of 18 healthy subjects. The concen-

tration of both cytokines was higher in patients with GD and

lower in patients with HT, compared to the control group

(p50.001). After treatment, the levels of cytokines TNF-aand IL-6 decreased in patients with GD (p50.01), but

increased in the group with HT (p50.05). In a similar study,

Akalin et al. [22] reported elevated levels of IL-6 in patients

with hyperthyroidism in the course of GD. The results of

these studies suggest that IL-6 may modulate the function of

the thyroid cells, because it plays an important role in B cell

differentiation and T cell activation. The release of IL-6 by

Table 5. Area under the curve: GD versus control group.

Asymptotic 95%confidence interval

Variable AreaStandard

errorAsymptoticsignificance

Lowerbound

Upperbound

IL-6 (pg/ml) 0.570 0.091 p¼ 0.435 (ns) 0.392 0.749TNF-a (pg/ml) 0.652 0.085 p¼ 0.092 (ns) 0.485 0.820IL-1b (pg/ml) 0.523 0.094 p¼ 0.801 (ns) 0.338 0.707

DOI: 10.3109/08916934.2014.914175 IL-1�, TNF-� and IL-6 in AITD in children 5

Aut

oim

mun

ity D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y T

ham

mas

at U

nive

rsity

on

10/0

8/14

For

pers

onal

use

onl

y.

thyrocytes may increase the intrathyroidal immune response

in Graves’ disease.

In the present study we found a positive correlation of IL-6

and TNF-a in hyperthyroidism (r¼ 0.62, p50.01). This

aspect has also been observed in the other authors’ studies

[21]. The increased TNF-a concentration is related to the

increasing serum IL-6 level in the hyperthyroid children.

Jiskra et al. [23] determined concentrations of cytokine

panel including sIL-2R, IFN-g, TNF-a, IL-10, CD30 and Bcl-

2 in 28 untreated patients with GD and 24 with HT, compared

to 15 healthy subjects, finding elevated TNF-a, IL-10, sIL-2R

in patients with GD compared to patients with HT and the

control group.

Diez et al. [13] studied 20 untreated patients with

hyperthyroidism and 20 with untreated hypothyroidism and

reported elevated levels of TNF-a and sTNFR-1 before

introducing treatment. After treatment of hyperthyroidism,

the authors achieved a significant reduction in TNF-a and

reduced levels of sTNFR-1, while L-thyroxine therapy in

patients with hypothyroidism did not bring changes in the

concentrations of both parameters. In another study, Aust

et al. [24] showed high levels of TNF-a mRNA in tissues of

patients with HT in contrast with low levels in patients with

GD, suggesting that thyroid cells derived from patients with

hypothyroidism may potentially release TNF-a.

The concentrations of TNF-a in our study did not differ

significantly between the studied groups. However, ROC

curve analysis showed that TNF-a may differentiate healthy

children and children with hypothyroidism (AUC¼ 0.691,

p¼ 0.034) with low sensitivity (54.5%) and high specificity

(85%) at a designated cut-off point of 14.7 pg/ml. Therefore,

TNF-a can be considered as a marker for hypothyroidism

in cAIT and its concentrations below a cut-off point of

14.7 pg/ml may be helpful in ruling-out cAIT in a given

patient. Moreover, children with hypothyroidism showed a

significant positive correlation between TNF-a and ATPO

(r¼ 0.54, p50.01). A similar correlation between ATPO and

secretion of cytokines TNF-a and IFN-g was presented in the

work of Karanikas et al. [25]. The hypothesis was recently

confirmed by demonstrating that ATPO antibodies may

promote thyroid peroxidase – dependent on the production

of cytokines, including IFN-g, TNF-a and IL-6 [26].

The role of cytokines in the pathogenesis of AITD is still

unclear, and the data in the literature on the clinical effects of

cytokines within the thyroid gland are controversial. To date,

only a few studies have demonstrated a relationship between

IL-6 and HT. In animal experiments, IL-6 plays a minor role

in the development of autoimmune thyroiditis [27]. In a recent

study, Rugerri et al. [28] found elevated levels of IL-6 in

patients with HT. Similar results were obtained by Sieminska

et al. [29]. The main findings of the study included

demonstration of high levels of IL-6 in the serum of patients

with HT, suggesting that IL-6 is involved in the development

of the disease in postmenopausal women.

In our study, no significant difference in IL-6 concentra-

tions between groups was found, which may indicate that IL-6

does not play a significant pathological role in the group of

young patients with AITD. Nonetheless, further studies are

needed to clarify this issue and to determine the potential

implications of serum IL-6 for diagnosis of AITD in children.

In conclusion, we demonstrated increased IL-1b levels in

children with cAIT. Our findings indicate that both IL-1b and

TNF-a in serum could be used as markers for differentiation

of children with cAIT from healthy children. Moreover,

significantly higher concentrations of IL-1b in children with

hypothyroidism may be used to distinguish between children

with cAIT and GD. Our results suggest that serum assays of

selected cytokines: IL-1b and TNF-a may provide valuable

clinical information relevant for the diagnosis and follow-up

of AITD in children.

Declaration of interest

The authors state that they have no conflicts of interest.

The study was funded from the research grant of the

Department (UM 502-01-01104118-06037).

References

1. Rasmussen, A. K., K. Bendtzen, and U. Feldt-Rasmussen. 2000.Thyrocyte-interleukin-1 interactions. Exp. Clin. Endocrinol.Diabetes. 108: 67–71.

2. Shi, Y., H. Wang, Z. Su, et al. 2010. Differentiation imbalance ofTh1/Th17 in peripheral blood mononuclear cells might contributeto pathogenesis of Hashimoto’s thyroiditis. Scand. J. Immunol. 72:250–255.

3. Bossowski, A., M. Moniuszko, M. Dabrowska, et al. 2013. Lowerproportions of CD4+CD25(high) and CD4+FoxP3, but notCD4+CD25+CD127(low) FoxP3+ T cell levels in children withautoimmune thyroid diseases. Autoimmunity. 46: 222–230.

4. Klatka, M., E. Grywalska, M. Partyka, et al. 2014. Th17 and Tregcells in adolescents with Graves’ disease. Impact of treatment withmethimazole on these cell subsets. Autoimmunity. 47: 201–211.

5. Luu, V. P., M. I. Vazquez, and A. Zlotnik. 2014. B cells participatein tolerance and autoimmunity through cytokine production.Autoimmunity. 47: 1–12.

6. Peng, D., B. Xu, Y. Wang, et al. 2013. A high frequency ofcirculating th22 and th17 cells in patients with new onset Graves’disease. PLoS One. 8: e68446.

7. Gianoukakis, A. G., and T. J. Smith. 2004. The role of cytokines inthe pathogenesis of endocrine disease. Can. J. Diabetes. 28: 30–42.

8. Rasmussen, A. K. 2000. Cytokine actions on the thyroid gland.Dan. Med. Bull. 47: 94–114.

9. Palczewska, I., and Z. Niedzwiedzka. 2001. Somatic developmentindices in children and youth of Warsaw. Med. Wieku. Rozwoj. 5:18–118.

10. Phenekos, C., A. Vryonidou, A. D. Gritzapis, et al. 2004. Th1and Th2 serum cytokine profiles characterize patients withHashimoto’s thyroiditis (Th1) and Graves’ disease (Th2).Neuroimmunomodulation. 11: 209–213.

11. Nanba, T., M. Watanabe, N. Inoue, and Y. Iwatani. 2009. Increasesof the Th1/Th2 cell ratio in severe Hashimoto’s disease and in theproportion of Th17 cells in intractable Graves’ disease. Thyroid. 19:495–501.

12. Rebuffat, S. A., M. Kammoun-Krichen, I. Charfeddine, et al. 2013.IL-1beta and TSH disturb thyroid epithelium integrity in auto-immune thyroid diseases. Immunobiology. 218: 285–291.

13. Diez, J. J., A. Hernanz, S. Medina, et al. 2002. Serum concentra-tions of tumour necrosis factor-alpha (TNF-alpha) and solubleTNF-alpha receptor p55 in patients with hypothyroidism andhyperthyroidism before and after normalization of thyroid function.Clin. Endocrinol. (Oxf). 57: 515–521.

14. Pedro, A. B., J. H. Romaldini, and K. Takei. 2011. Changesof serum cytokines in hyperthyroid Graves’ disease patientsat diagnosis and during methimazole treatment.Neuroimmunomodulation. 18: 45–51.

15. Salvi, M., M. Pedrazzoni, G. Girasole, et al. 2000. Serumconcentrations of proinflammatory cytokines in Graves’ disease:effect of treatment, thyroid function, ophthalmopathy and cigarettesmoking. Eur. J. Endocrinol. 143: 197–202.

6 H. Mikos et al. Autoimmunity, Early Online: 1–7

Aut

oim

mun

ity D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y T

ham

mas

at U

nive

rsity

on

10/0

8/14

For

pers

onal

use

onl

y.

16. Lakatos, P., J. Foldes, C. Horvath, et al. 1997. Serum interleukin-6and bone metabolism in patients with thyroid function disorders.J. Clin. Endocrinol. Metab. 82: 78–81.

17. Celik, I., S. Akalin, and T. Erbas. 1995. Serum levels of interleukin6 and tumor necrosis factor-alpha in hyperthyroid patients beforeand after propylthiouracil treatment. Eur. J. Endocrinol. 132:668–672.

18. Siddiqi, A., J. P. Monson, D. F. Wood, et al. 1999. Serum cytokinesin thyrotoxicosis. J. Clin. Endocrinol. Metab. 84: 435–439.

19. Wakelkamp, I. M., M. N. Gerding, J. W. Van Der Meer, et al.2000. Both Th1- and Th2-derived cytokines in serum areelevated in Graves’ ophthalmopathy. Clin. Exp. Immunol. 121:453–457.

20. Pichler, R., W. Maschek, M. Hatzl-Griesenhofer, et al. 2003.Soluble tumour necrosis factor-alpha receptor I and interleukin-6 asmarkers of activity in thyrotoxic Graves’ disease. Horm. Metab.Res. 35: 427–433.

21. Kiziltunc, A., M. Basoglu, B. Avci, and I. Capoglu. 1999. SerumIL-6 and TNF-a in patients with thyroid disorders. Turk. J. Med.Sci. 29: 25–30.

22. Akalin, A., O. Colak, O. Alatas, and B. Efe. 2002. Boneremodelling markers and serum cytokines in patients with hyper-thyroidism. Clin. Endocrinol. (Oxf). 57: 125–129.

23. Jiskra, J., M. Antosova, Z. Limanova, et al. 2009. The relationshipbetween thyroid function, serum monokine induced by interferon

gamma and soluble interleukin-2 receptor in thyroid autoimmunediseases. Clin. Exp. Immunol. 156: 211–216.

24. Aust, G., and W. A. Scherbaum. 1996. Expression of cytokines inthe thyroid: thyrocytes as potential cytokine producers. Exp. Clin.Endocrinol. Diabetes. 104: 64–67.

25. Karanikas, G., M. Schuetz, K. Wahl, et al. 2005. Relation of anti-TPO autoantibody titre and T-lymphocyte cytokine productionpatterns in Hashimoto’s thyroiditis. Clin. Endocrinol. (Oxf). 63:191–196.

26. Nielsen, C. H., T. H. Brix, R. G. Leslie, and L. Hegedus. 2009. Arole for autoantibodies in enhancement of pro-inflammatorycytokine responses to a self-antigen, thyroid peroxidase. Clin.Immunol. 133: 218–227.

27. Mori, K., K. Yoshida, M. Mihara, et al. 2009. Effects of interleukin-6 blockade on the development of autoimmune thyroiditis innonobese diabetic mice. Autoimmunity. 42: 228–234.

28. Ruggeri, R. M., S. Sciacchitano, A. Vitale, et al. 2009. Serumhepatocyte growth factor is increased in Hashimoto’s thyroiditiswhether or not it is associated with nodular goiter as compared withhealthy non-goitrous individuals. J. Endocrinol. Invest. 32:465–469.

29. Sieminska, L., C. Wojciechowska, B. Kos-Kudla, et al. 2010.Serum concentrations of leptin, adiponectin, and interleukin-6 inpostmenopausal women with Hashimoto’s thyroiditis. Endokrynol.Pol. 61: 112–116.

DOI: 10.3109/08916934.2014.914175 IL-1�, TNF-� and IL-6 in AITD in children 7

Aut

oim

mun

ity D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y T

ham

mas

at U

nive

rsity

on

10/0

8/14

For

pers

onal

use

onl

y.