Chronic autoimmune thyroiditis in children and adolescents: at presentation and during long-term follow-up
- Institute for Endocrinology and Diabetes, National Center for Childhood Diabetes, Schneider Children’s Medical Center of Israel, Petah Tikva, and Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel
- Professor M Phillip, Institute for Endocrinology and Diabetes, National Center for Childhood Diabetes, Schneider Children’s Medical Center of Israel, Petah Tiqwa 49202, Israel;
- Accepted 18 July 2008
- Published Online First 14 August 2008
Objective: To investigate the clinical manifestations of autoimmune thyroiditis (AIT) leading to referral in children and adolescents, in addition to disease course and long-term outcome.
Design: Chart review.
Setting: Major tertiary hospital.
Patients: 114 children/adolescents (92 female, 22 male; mean (SD) age 11.8 (35) years) with AIT referred for evaluation/treatment.
Main outcome measures: Clinical characteristics at presentation, reasons for referral, treatment and long-term (mean 6 years) outcome; by thyroid and pubertal status.
Results: The male/female (1:4.2) ratio was lower than in adult AIT (1:10) and varied by age. Patients with noticeable goitre at presentation (39.5%) accounted for half the total number in whom goitre was eventually diagnosed. Other reasons for referral were clinical symptoms of hypothyroidism (28.9%) and findings on work-up for an unrelated problem (19.2%) or for high-risk groups (10.5%). There was no difference in management or outcome between patients who underwent ultrasound (n = 79) or not. Treatment was initiated shortly after diagnosis in all 42 hypothyroid patients and 44/48 compensated hypothyroid patients, and within 16 months in 19/24 euthyroid patients. There was no change in thyroid status in the nine untreated patients. Height standard deviation score (SDS) was normal at referral and last visit and correlated with parental height SDS. Puberty was normal. There was no significant difference in body mass index SDS at referral by pubertal or thyroid status. There was no difference from the general population in the prevalence of obesity.
Conclusions: Although goitre is the main symptom leading to diagnosis of AIT, it is still often overlooked, underscoring the need for thorough thyroid evaluation on routine physical examination. Acquired hypothyroidism is not often associated with obesity, and ultrasound usually has no added diagnostic value. Adequate treatment in this age group leads to normal growth, puberty and final height.
Despite the relatively common prevalence of autoimmune thyroiditis (AIT),1 the paediatric literature on the clinical presentation remains sparse. This is important because a prolonged hypothyroid state before treatment in children may result in incomplete catch-up growth,2 obesity3 and precocious pseudo-puberty.4
What is already known on this topic
Autoimmune thyroiditis (AIT) is the most common acquired thyroid disease in children and adolescents in the western world.
Manifestations of acquired hypothyroidism in children include growth retardation and/or delayed puberty, in association with bone age retardation, obesity and, in rare cases, precocious puberty.
What this study adds
Although goitre is the main symptom leading to a diagnosis of AIT, it is still often overlooked, underscoring the need for thorough thyroid evaluation on routine physical examination.
Thyroid ultrasound scan, often performed in children with AIT, does not change the treatment, clinical course or outcome, and thus should not be indicated routinely.
Furthermore, findings vary, because some studies investigated patients detected during screening of healthy populations,5–8 whereas others were limited to selected patients with goitre.9 10 Also, most studies failed to describe the clinical picture leading to referral, but rather focused on natural history5 9 11 12 or dynamics of the thyromegaly7 13 or growth.2
The present study aimed at investigating the clinical manifestations of AIT leading to referral in children and adolescents from an iodine-sufficient region. We also examined the disease course and long-term outcome.
A diagnosis registry search of our university-affiliated tertiary centre yielded 114 children and adolescents with AIT who were referred for diagnosis and treatment between 1987 and 2003 and were followed for at least 12 months. Patients with diabetes mellitus were excluded. The reasons for referral were derived from the charts. The diagnosis of AIT was based on findings of seropositivity for thyroglobulin autoantibodies and/or thyroid peroxidase autoantibodies, accompanied by at least one of the following: abnormal thyroid function; enlarged thyroid gland; morphological changes on thyroid ultrasound. Thyroid gland enlargement was assessed by palpation according to the WHO criteria.14
Basal concentrations of thyroid-stimulating hormone (TSH) (normal reference 0.4–4 mIU/l) and free thyroxine (FT4; normal 10.5–25.7 pmol/l) were analysed with a commercial chemiluminescent enzyme immunoassay (Diagnostic Products Corp, Los Angeles, California, USA). Anti-(thyroid peroxidase) (normal reference <75 IU/ml) and anti-thyroglobulin IgG (normal <150 IU/ml) were initially measured by the haemagglutination method and, more recently, by enzyme-linked immunosorbent assay (Orgentec Diagnostika, Maintz, Germany). On the basis of the initial thyroid function studies, patients were categorised as euthyroid (normal-range FT4 and TSH), compensated hypothyroid (normal FT4, TSH >4.0) or hypothyroid (FT 4<10.5, TSH >4.0).
Imaging studies, performed in some patients, included ultrasound and thyroid scanning with technetium.
Treatment consisted of l-thyroxine (LT4). FT4 and TSH measurements were repeated every 6 months, and the dose of LT4 was adjusted to keep the concentrations within normal range. None of the patients took drugs known to affect thyroid function.
Physical examination included measurements of height (with the Harpenden–Holtain stadiometer) and weight. Height was calculated as height standard deviation score (Ht-SDS) for the patients and for both their parents using the Centers for Disease Control growth charts.15 Body weight was expressed as body mass index (BMI), and BMI-SDS was calculated.16 Pubertal status was characterised by Tanner stage as follows17 18: prepubertal, Tanner 1; pubertal, Tanner 2–4; postpubertal, Tanner 5. Bone age was estimated as described by Greulich and Pyle.19
Patients were also investigated for other autoimmune disorders and a family history (first-order or second-degree relative) of autoimmune disease, with emphasis on thyroid disorders.
The institutional human research committee approved the study.
Statistical analyses were performed with the BMDP program. The results are given as mean (SD). One-way analysis of variance was used to compare clinical, anthropometric and laboratory parameters by thyroid status; two-way analysis of variance was used to compare variables among the pubertal subgroups. Multiple comparisons were made using Bonferroni corrections. Spearman rank correlation was used to evaluate correlations between various parameters. p⩽0.05 was considered significant.
Clinical characteristics at referral
The sample included 92 female and 22 male patients (4.2:1) aged 1–19 years (table 1). In patients older than 10 years of age, the male/female ratio was significantly lower (1:3.1) than in younger patients (1:14.5) and adults (1:10 to 1:20).20
The most common reason for referral was thyroid gland enlargement (39.5%), followed by clinical symptoms of hypothyroidism (28.9%) (table 1). In the remaining patients, thyroid dysfunction was found on work-up performed for an unrelated medical problem or routine screening of high-risk groups
Twenty-four patients (21%) were euthyroid, 48 (42%) had compensated hypothyroidism, and 42 (37%) had hypothyroidism (including two patients in whom hyperthyroidism at presentation reversed to hypothyroidism within weeks) (table 2). Physical examination revealed goitre in 88 patients: 4/22 boys and 74/92 girls. The prevalence was significantly higher in the euthyroid group (92%) than in the other groups (p = 0.04). There was no difference in clinical symptoms of hypothyroidism by thyroid status, except for a higher rate of constipation in the hypothyroid group (table 2).
No significant difference was found in BMI-SDS or in prevalence of BMI-SDS >1.0 at referral by pubertal group or thyroid status. Five of the 20 patients (25%) with BMI-SDS >1.0 had hypothyroidism compared with 36 of the 94 patients (38%) with BMI-SDS <1.0.
Fifty-nine patients (52%) had a family history of AIT (table 1). No differences were found between patients with or without a family history of AIT in any of the clinical, laboratory or anthropometric parameters. Euthyroid patients with a family history had a lower likelihood of remaining euthyroid than those without, but the difference was not significant (12.5% vs 50%, p = 0.125).
Laboratory findings and imaging
The prevalence of initial laboratory findings compatible with hypothyroidism did not differ significantly by reason for referral (table 3).
Ultrasound scan was requested by the referring doctor for 58 patients and by the endocrinologist for 21 (for a finding of a nodule, multinodularity or asymmetry of the gland on palpation in 18, and unknown reason in three). Thyroid scintigraphy was requested by the referring doctor for 24 patients and by the endocrinologist for 18 (for hyperthyroidism in two and a “suspected” nodule in 16). Scintigraphy findings were unremarkable, apart from thyroid enlargement. Findings in all five patients who underwent fine-needle aspiration of suspected nodules were benign and consistent with chronic lymphocytic thyroiditis. No significant differences were found in laboratory or anthropometric data at referral or at last visit between patients who had or did not have an ultrasound scan.
Growth and puberty
Thirty-two patients were prepubertal at diagnosis (mean age 8.0 (4.6) years), 41 were pubertal (mean age 11.6 (2.1) years), and 41 were postpubertal (mean age 14.9 (1.8) years). Clinical, laboratory and anthropometric parameters were not significantly different between the pubertal groups.
Both onset and duration of puberty were normal in the children (male and female) who were prepubertal at diagnosis. Girls in this subgroup had menarche at a mean age of 12.5 (0.5) years.
To avoid skewing the results, three girls with Turner syndrome, two with Down syndrome and two who presented with hyperthyroidism were excluded from the group analysis of Ht-SDS. The mean duration of follow-up was 6.0 years. Ht-SDS at last follow-up was significantly lower than at diagnosis for the entire study group (−0.45 (1.13) vs −0.05 (1.45), p<0.001) and for the pubertal groups. No difference was found in Ht-SDS at referral or at last visit by pubertal status at diagnosis. The first Ht-SDS was significantly higher in the euthyroid than the hypothyroid patients (both compensated and overt), but the difference was not maintained at the last follow-up.
The first Ht-SDS was significantly lower in patients referred for clinical symptoms of hypothyroidism (−0.45 (1.18)) than in patients referred for goitre (0.40 (1.51)) (p<0.05). No significant difference was noted between patients referred for goitre and patients identified incidentally (−0.02 (1.21)) or on routine examination 0.34 (0.48)).
Last Ht-SDS correlated with both paternal Ht-SDS (R = 0.40, p<0.001) and maternal Ht-SDS (R = 0.37, p<0.001). Ninety-five patients (83%) had reached their final height by the last follow-up visit (mean Ht-SDS −0.41(1.16)).
LT4 therapy was initiated in all 42 hypothyroid patients. Of the 48 patients with compensated hypothyroidism, 24 had a TSH concentration >10.0 mIU/l, including 20 who started treatment at diagnosis and four (TSH 10–23.7 mIU/l) who started treatment 35–174 weeks after diagnosis. The other 24 patients in this group had TSH concentrations below 10.0 mIU/l: 20 were treated for the relative indication of goitre associated with raised TSH, and four were not treated and remained compensated hypothyroid during 2–5.8 years of follow-up (fig 1).
Nineteen of the 24 euthyroid patients were also treated with LT4 within a mean of 16 months (3–36) after diagnosis. Indications for treatment were further thyroid enlargement during follow-up, significant increase in TSH concentration or decrease in FT4 concentration, or both, or one or more symptoms compatible with hypothyroidism. The other five euthyroid patients remained euthyroid without treatment during follow-up of 2–16.6 years. Compliance was good in all treated patients, as indicated by the normal-range concentrations of TSH and FT4 at most visits and direct questioning (fig 1).
The mean initial dose of LT4 for the whole sample was 1.5 (0.8) μg/kg/day (range 0.3–5). There were no significant differences in dose during follow-up by thyroid status or reason for referral (fig 1).
We investigated the clinical presentation and long-term outcome of children and adolescents with AIT who were referred to a tertiary care centre. The most common reason for referral was goitre, in 39% of patients; however, its actual prevalence on examination was almost twofold higher (77%). The finding that almost half the cases of goitre were noted only when the patients were investigated for other reasons underscores the need for thorough thyroid examination as part of the routine physical examination in apparently healthy children.
In almost one-third of patients, AIT was diagnosed in the absence of clinical symptoms, either incidentally or on routine screening for high-risk groups. Interestingly, the patients considered to be high risk showed no differences in anthropometric, clinical or laboratory data from the patients referred for goitre. More surprising was the same prevalence of overt hypothyroidism in this group as in patients referred for goitre or clinical symptoms of hypothyroidism. Thus, periodic thyroid function tests may be indicated in people at high risk of AIT.21–25
The reasons for the female predilection for AIT (4.2:1 in this study), especially in children younger than 10 years, are unknown. They may be related to chromosomal factors, such as skewed X-chromosome inactivation,26 or to female sex steroids.27 The rise in oestrogen concentrations in boys during puberty may be a contributing factor to the increased number of boys with AIT in the pubertal group.
Our 52% rate of positive family history supports previous findings of a familial clustering of AIT disease based on thyroid antibody and ultrasound studies28 or fine-needle aspiration cytology.23 By relying on medical history, we may have missed some undiagnosed familial cases and underestimated the actual rate.
Most patients who underwent thyroid ultrasound scan were referred by their paediatrician as part of routine work-up; 21 were referred by our team for a “suspicious” finding on palpation. Given the negligible effect of the ultrasound findings on either treatment or outcome in our sample, we suggest that ultrasound should not be routinely performed in children with laboratory-confirmed AIT in the absence of another clinical indication.
Thyroid scintigraphy, too, was usually normal and non-contributory, as previously demonstrated.29
Mean Ht-SDS at presentation was within the normal range. The lower Ht-SDS at the last follow-up than at referral in all thyroid groups, despite the good compliance and absence of evidence of over-treatment, may at least partially be explained by the correlation of the last Ht-SDS with both maternal and paternal Ht-SDS. Although the difference in mean Ht-SDS between these time points reached significance (p<0.001), both values were within the normal range. Thus, although final height may be slightly compromised in acquired hypothyroidism, it remains within both the normal range and the expected genetic height range.
Our finding that neither pubertal stage nor thyroid function at initiation of treatment influenced the last Ht-SDS agrees with the study of Jaruratanasirikul et al.7 However, Rivkees et al2 reported that a selected group of patients with severe long-standing hypothyroidism associated with severe growth retardation failed to achieve normal adult stature. In our sample, the normal growth and puberty in the hypothyroid group may be attributable to the short duration of the hypothyroidism, owing to early detection and appropriate treatment.
The similar prevalence of obesity or overweight to that of the general population30 indicates that acquired hypothyroidism in children is often not associated with obesity.
Thus, both growth retardation and obesity, usually considered clinical hallmarks of acquired hypothyroidism in childhood, were uncommon findings in our study group.
The decision to treat was made over the years by different doctors according to clinical judgment. Most patients with compensated hypothyroidism, and also euthyroid patients, were treated for thyroid enlargement, a further increase in TSH concentration, or clinical symptoms of hypothyroidism. In most euthyroid patients, treatment was initiated within a median of 1 year of presentation. Throughout follow-up, results of the thyroid function tests remained within the normal range, in association with normal growth velocity and puberty. Regular follow-up eliminated the risk of over-treatment. At the same time, studies have shown that treating subclinical hypothyroidism may confer benefits, such as inhibiting progression to overt hypothyroidism,31 32 reducing thyroid size,13 reducing the risk of coronary heart disease33 and muscle dysfunction,34 improving he lipid profile,35 reversing intraocular pressure associated with subclinical hypothyroidism,36 and improving growth velocity in children and adolescents.37
The high rate of goitre detected in our series of children and adolescents with AIT underscores the need for a thorough thyroid evaluation as part of the routine screening.
Acquired hypothyroidism is not often associated with obesity.
Ultrasound has no diagnostic value in AIT.
Adequate treatment of children with AIT leads to normal growth and puberty and normal final height, irrespective of pubertal stage or thyroid function at diagnosis.
Although Ht-SDS may be slightly compromised, it remains within the normal range during the clinical course of AIT and correlates with parental height.
LT4 treatment of euthyroid children with AIT, if appropriately monitored, does not seem to be harmful and may be beneficial.