- Email: [email protected]
Ultrasound Appearance of Thyroid Tissue in Hypothyroid Infants
Ultrasound Appearance of Thyroid Tissue in Hypothyroid Infants EISHIN OGAWA, MD, KANAKO KOJIMA-ISHII, MD,
AND IKUMA
FUJIWARA, MD
Objective To identify those infants who need a higher starting dose of levothyroxine (l-T4) for early normalization of thyroid-stimulating hormone (TSH) level. Study design TSH levels at 2 time points (1 to 3 weeks and 3 to 5 weeks) after l-T4 therapy at a starting dose of 8 to12 g/kg/day were evaluated retrospectively in 22 hypothyroid infants screened for congenital hypothyroidism (CH) in terms of etiology as determined by ultrasonography (US), the size of distal femoral epiphysis (DFE), and initial thyroid function. Results The infants with a noneutopic thyroid or small DFE exhibited significantly higher posttherapeutic TSH levels compared with the other infants. Eight of the 9 infants who failed to achieve normalized TSH values at 1 to 3 weeks had noneutopic thyroid. All of the infants with eutopic thyroid exhibited normalized TSH at 3 to 5 weeks, and a significantly greater proportion of the infants with eutopic thyroid exhibited normalized TSH at 1 to 3 weeks compared with those with noneutopic thyroid. Stepwise regression analysis demonstrated that US etiology was a significant independent variable for normalization of TSH at 1 to 3 weeks. Conclusions US examination to identify eutopic or noneutopic thyroid provides useful information for determining the starting dose of l-T4 in hypothyroid infants. (J Pediatr 2008;153:101-4) lthough the introduction of newborn screening has greatly improved the intellectual outcome in congenital hypothyroidism (CH), closing the remaining subtle neurologic gap in children with severe CH remains a work in progress.1,2 Using a higher starting dose to more quickly normalize thyroid-stimulating hormone (TSH) levels could result in normalization of developmental IQ even in those with severe CH.3-6 Consequently, an initial levothyroxine (l-T4) dose of 10 to 15 g/kg/day is now recommended.7-9 However, it is not clear whether replacement therapy should start with 10 or 15 g/kg/day; the recommendation is vague, specifying only “depending on the severity of the initial hypothyroidism”7 or “in the severe cases.”9 Furthermore, although it has been shown that an initial l-T4 dose of 12 to 17 g/kg/day can raise the TSH level to the target range in 2 weeks,10 all hypothyroid infants might not need the full dose. In CH, disease severity is indicated by etiology and skeletal maturity, in addition to the degree of initial hyperthyroidism. Because infants with athyreosis exhibit delayed skeletal maturity11 and lower initial free thyroxine (fT4) levels12 and take longer to achieve normal TSH values13,14 with more frequent dosage adjustment14 compared with infants with ectopic thyroid or dyshormonogenesis, having a different management strategy for each etiology is clearly the ideal situation. To obtain a precise diagnosis, 123I-scintigraphy (but not technetium scanning) has been recommended.7,15,16 However in Japan, as a practical matter, performing 123I-scintigraphy during the early neonatal period is difficult without delaying the start of treatment; thus, it is often performed later (eg, at age 5 years). Thyroid ultrasonography (US), an alternative approach, is not considered to give additional information useful for management, probably due to its weakness in detecting the ectopic gland, the most common cause of CH. Here we report the different responses to initial treatment in infants with CH depending on their US-determined etiology.
A
METHODS Screening Procedure In Japan, neonatal CH screening is based primarily on the measurement of TSH. Any infant with a TSH value ⱖ30 mU/L in a dried blood sample on filter paper at age
CH DFE fT4
Congenital hypothyroidism Distal femoral epiphysis Free thyroxine
l-T4 TSH US
Levothyroxine Thyroid-stimulating hormone Ultrasonography
From the Department of Pediatrics, Tohoku University School of Medicine, Sendai, Japan. Submitted for publication May 23, 2007; last revision received Oct 31, 2007; accepted Dec 5, 2007. Reprint requests: Eishin Ogawa, 2-1 SeiryoMachi, Aoba-ku, Sendai 980-8575, Japan. E-mail: [email protected]. 0022-3476/$ - see front matter Copyright © 2008 Mosby Inc. All rights reserved. 10.1016/j.jpeds.2007.12.015
101
5 days is immediately referred to a pediatrician for diagnostic evaluation. Any infant with a borderline result (10 ⱕ TSH ⬍ 30 mU/L) undergoes a second heel puncture and is referred if the result is again ⱖ10 mU/L.
Subjects A total of 195,621 neonates were screened for CH in Miyagi Prefecture, Japan, during the 9-year period between April 1995 and March 2004. The referral rate during this period was 0.22%. A total of 386 neonates (90% of those with positive results) were referred to Tohoku University Hospital. Of these, 130 infants were treated with l-T4 either transiently or permanently: 59 with mild CH (pretreatment serum TSH ⬍20 mU/L), 40 with moderate CH (pretreatment serum TSH 20 to 80 mU/L), and 31 with severe CH (pretreatment TSH ⬎80 mU/L). Only those with severe CH were candidates for this study, because the study’s major objective was to identify those infants who should require a higher dose (eg, 15 g/kg/day). Those infants referred after the second heel puncture were excluded, because their hypothyroidism could be affected by environmental factors after birth (eg, iodine exposure). Thus, the enrollment criteria were as follows: (1) abnormal initial screening result (dried blood TSH ⱖ30 mU/L); (2) confirmatory serum TSH ⬎80 mU/L and fT4 ⬍1.2 ng/dL; and (3) an l-T4 starting dose of ⱖ8 g/kg/day. A total of 67 infants exhibited abnormal initial results; of these, 25 were selected based on criterion 2, and 3 were excluded based on criterion 3. One of these latter infants could not tolerate oral l-T4 due to gastrointestinal malformation, and the other 2 were treated with lower doses because of suspected transient CH born to mothers with Graves’ disease taking antithyroid medication. Thus, a total of 22 infants were evaluated. These infants were allocated according to factors indicating disease severity, such as confirmatory fT4 and TSH levels, size of the distal femoral epiphysis (DFE), and US-derived etiology, as described later. Ultrasound Scanning A real-time mechanical sector scanner (EUB-8000 ultrasound scanner; Hitachi, Tokyo, Japan) with a digital beamformer system and a 10-MHz probe was used for the thyroid scan. A thyroid gland clearly visible at the normal site on US was classified as eutopic. In our previous 2-year survey in 97 infants with positive screening results, 88 had eutopic thyroid (5 as enlarged thyroid), 7 had noneutopic thyroid, and 2 had hemiagenesis.17 Eutopic thyroid suggests normal thyroid or dyshormonogenesis as a cause of CH, and noneutopic thyroid suggests dysgenesis, such as occurs in athyreotic, hypoplastic, or ectopic thyroid. Of the 22 infants in the present study, 8 had eutopic thyroid and 14 had noneutopic thyroid. A total of 16 infants with noneutopic thyroid were identified (about 1 in 11,000) during the 9-year study period; 15 of these demonstrated initial abnormal screening results, 14 of whom were enrolled in the present study. 102
Ogawa, Kojima-Ishii, and Fujiwara
Initial Treatment and Management Replacement therapy was started promptly (on day 15.5 ⫾ 6.7 on average) with an initial dose of 8 to 12 g/kg/day (mean, 9.66 ⫾ 0.98 g/kg/day) after blood sampling, X-ray to evaluate skeletal maturity, and thyroid US. TSH and fT4 measurements were obtained between 1 and 3 weeks and again between 3 and 5 weeks after the treatment. The l-T4 dose was increased in 6 infants but decreased in none, based on the TSH and fT4 values at 1 to 3 weeks. The exact values of TSH and fT4, along with the numbers of infants whose TSH levels reached within the target range, were compared among the groups. All infants with mild or moderate CH (serum TSH ⬍80 mU/L), who were excluded from this evaluation, were treated with ⱕ10 g/kg/day, and all reached the target TSH range at 3 to 5 weeks. Skeletal Maturity Skeletal maturity was assessed based on the size of the DFE, defined as the ratio of the transverse diameter between the distal epiphysis and the metaphysis of the femoral bones. The mean DFE size in term Japanese newborns is 0.31 ⫾ 0.07118; a value ⬍⫺2 standard deviations (0.168) indicates a small DFE. A small DFE was found in 6 of the 14 infants with noneutopic thyroid and in none of the 8 infants with eutopic thyroid. Hormone Measurements Serum TSH and fT4 levels were measured by sandwich enzyme imuumoassay (AxSYM; Dainabot, Tokyo, Japan). The normal values for infants in our hospital are as follows: TSH, 0.3 to 10.0 mU/L at age 1 month and 0.3 to 7.5 mU/L at age 2 months; fT4, 1.2 to 2.0 ng/dL for the entire period. Serum fT4 levels at confirmation ⬍0.5 ng/dL were designated “very low” (n ⫽ 12), and those ⱖ0.5 ng/dL (but ⬍1.2 ng/dL) were designated “low.” Very low fT4 was found in 5 of the 8 infants with eutopic thyroid and in 7 of the 14 infants with noneutopic thyroid. Serum TSH levels at confirmation ⬎300 mU/L were designated “very high” (n ⫽ 12), and those ⱕ300 mU/L (but ⬎80 mIU/L) were designated “high” (n ⫽ 10). Very high TSH was found in 2 of the 8 infants with eutopic thyroid and in 8 of the 14 infants with noneutopic thyroid. The target range for TSH is ⬍10 mU/L at 1 to 3 weeks and ⬍7.5 mU/L at 3 to 5 weeks. Statistical Analysis The differences in serum levels of TSH and fT4 at 3 time points in the various categories of patients were analyzed using the nonparametric Mann-Whitney U test. The distribution of infants achieving the target TSH range at 2 time points after l-T4 therapy was compared among the groups using 2 tests. Stepwise regression analysis was performed to identify the variables contributing to the definition of normalized TSH. Statistical significance was accepted at P ⬍ .05. The Journal of Pediatrics • July 2008
Figure 1. TSH levels (mU/L) at confirmation (A), 1 to 3 weeks (B), and 3 to 5 weeks (C). The data are medians, 25th and 75th percentiles, and ranges. Asterisks denote statistical significance (*P ⬍ .05; **P ⬍ .01; ***P ⬍ .001). The numbers in parentheses are the patient numbers.
RESULTS The serum TSH levels for each patient group are shown in Figure 1. Concentrations of serum confirmation TSH and fT4 levels differed significantly between the groups classified in terms of TSH and fT4 values, but not between the groups classified according to US-derived etiology and DFE size. The postreatment serum fT4 levels did not differ between the groups in each category (data not shown). The posttreatment serum TSH levels differed significantly between the groups categorized based on US-derived etiology and DFE size. TSH levels at 1 to 3 weeks and 3 to 5 weeks were higher in the noneutopic and small DFE groups compared with the eutopic and normal DFE groups. Thirteen infants (59.1%) reached the target TSH range at 1 to 3 weeks, and 17 (77.3%) did so at 3 to 5 weeks. The rates of TSH normalization for the groups of each category are shown in Figure 2. According to the US-derived etiology (but not the other categories), the distribution of infants with normalized TSH differed significantly at 1 to 3 weeks (2 ⫽ 4.197; P ⬍ .05), reflecting the greater proportion of infants with eutopic thyroid reaching the target range of TSH. In the eutopic group, 7 of 8 infants achieved normalized TSH at 1 to 3 weeks; the remaining infant, who had an abnormal value at 1 week, reached the target range at 4 weeks posttreatment without a dosage increment. In contrast, more than 50% of the infants with noneutopic thyroid failed to normalize TSH at 1 to 3 weeks, and 5 of these infants exhibited abnormal values at 3 to 5 weeks despite a dosage increment. A similar trend was observed in the other categories, with those infants with a small DFE, very low fT4, or very high TSH exhibiting lower rate of TSH normalization; none of these differences were statistically significant, however. Because those infants with noneutopic thyroid tended to have a small DFE (0.171 ⫾ 0.011 vs 0.241 ⫾ 0.005; P ⫽ Ultrasound Appearance of Thyroid Tissue in Hypothyroid Infants
Figure 2. Proportion of the infants who reached the target range of TSH levels (normalization rate) at 1 to 3 weeks (A) and 3 to 5 weeks (B). The white bars indicate children with normalized TSH; the solid bars indicate those with normalization failure of TSH. An asterisk (*) denotes statistical significance (P ⬍ .05). The numbers in parentheses are the patient numbers.
.1150) and more severe initial CH (baseline TSH higher in the noneutopic group, albeit not significantly so; see Figure 1), these categories are associated with one another. This finding led us to perform stepwise regression analysis to identify the significant variables defining normalization of TSH. Only the US-derived etiology was found to be a significant independent variable for normalization of TSH at 1 to 3 weeks (P ⬍ .05) and a nearly significant independent variable at 3 to 5 weeks (P ⫽ .058).
DISCUSSION In the present study, l-T4 therapy at a mean dose of 9.66 g/kg/day failed to normalize TSH levels at 3 to 5 weeks posttreatment in approximately 25% of the infants. Because the current goal of initial treatment of CH is to normalize TSH by age 1 month,7,8 in an effort to achieve good intellectual outcome even in the most severely affected infants, the higher dose was indicated for those who did not respond to the initial therapy, and characterization of the nonresponders was the study’s major aim. Responses to treatment differ dependent on the etiology, with infants with athyreosis requiring longer to achieve normalized TSH than those with ectopic thyroid and dyshormonogenesis.13,14 Therefore, we expected that the nonresponders would be demonstrated to have noneutopic thyroid by US. Indeed, 8 of the 9 nonresponders at 1 to 3 weeks and all of the nonresponders at 3 to 5 weeks had noneutopic thyroid. In addition to the fact that those infants with eutopic thyroid exhibited prompt normalization of TSH (Figures 1 and 2), and that stepwise regression analysis demonstrated that US-derived etiology is a significant factor in normalizing TSH, US also helped identify those infants requiring the higher l-T4 dosage. Although determining the etiology by 123I-scintigraphy is certainly ideal, US is an alternative technique when 103
123
I-scintigraphy is not available. US can accurately detect the thyroid gland in its normal location, but may fail to detect an ectopic gland,19 although a recent study has demonstrated markedly improved sensitivity of color Doppler US in diagnosing ectopic thyroid tissue.20 This weakness is probably the major reason why US has not been emphasized to date. Despite this drawback, however, the present study has demonstrated that a rough estimation with US, whether or not the thyroid is located normally, can provide useful information when determining the starting dose of l-T4. Furthermore, because it can be assumed that a small, ectopic thyroid with severe initial hyperthyroidism will clinically resemble athyreosis in terms of disease severity, the ability to differentiate a small ectopic thyroid from athyreosis should not be necessary for initial management planning. We propose providing a rough estimation with US before starting therapy, followed by subsequent precise diagnosis with 123I-scintigraphy, as is practical. DFE size, a measurement of skeletal maturity at birth, is considered to reflect the severity of CH in utero, with a caveat that because DFE depends on gestational age, a small DFE may be found in infants of shortened gestational age. The greater proportion of infants with athyreosis exhibit delayed skeletal maturation, and those with delayed skeletal maturation have intellectual impairment as well.11 Thus, DFE size is a factor in determining disease severity; however in our subjects, only those with noneutopic thyroid exhibited a small DFE. Although none of our infants with eutopic thyroid had a small DFE despite severe initial hyperthyroidism, infants with total dyshormonogenesis may exhibit a small DFE, and some infants with eutopic thyroid may need the maximum initial dose. A combination of US-derived etiology and DFE size might be advantageous for evaluating a nonresponder; such an evaluation was not feasible in our study due to the small number of subjects. In conclusion, although our results are preliminary (because of the small number of subjects and retrospective manner of the study design), they indicate that US to identify eutopic or noneutopic thyroid can be useful in determining the optimum starting dose of l-T4 for CH. When the USbased diagnosis is eutopic thyroid, an initial dose of approximately 10 g/kg/day appears to be sufficient (except possibly in those infants with a small DFE). In contrast, when the US-based diagnosis is noneutopic thyroid, an initial l-T4 dose
104
Ogawa, Kojima-Ishii, and Fujiwara
of up to 15 g/kg/day is advisable, to quickly normalize TSH levels.
REFERENCES 1. Hindmarsh PC. Optimisation of thyroxine dose in congenital hypothyroidism. Arch Dis Child 2002;86:73-5. 2. Rovet JF. In search of the optimal therapy for congenital hypothyroidism. J Pediatr 2004;144:698-700. 3. Dubuis JM, Glorieux J, Richer F, Deal CL, Dussault JH, van Vliet G. Outcome of severe congenital hypothyroidism: closing the developmental gap with early high-dose levothyroxine treatment. J Clin Endocrinol Metab 1996;81:222-7. 4. Bongers-Schokking JJ, Koot HM, Wiersma D, Verkerk PH, de Muinck KeizerSchrama SM. Influence of timing and dose of thyroid hormone replacement on development in infants with congenital hypothyroidism. J Pediatr 2000;136:292-7. 5. Selva KA, Harper A, Downs A, Blasco PA, LaFranchi SH. Neurodevelopmental outcomes in congenital hypothyroidism: comparison of initial T4 dose and time to reach target T4 and TSH. J Pediatr 2005;147:775-80. 6. Simoneau-Roy J, Marti S, Deal C, Huot C, Robaey P, Van Vliet G. Cognition and behavior at school entry in children with congenital hypothyroidism treated early with high-dose levothyroxine. J Pediatr 2004;144:747-52. 7. American Academy of Pediatrics. Update of newborn screening and therapy for congenital hypothyroidism. Pediatrics 2006;117:2290-303. 8. Toublanc JE on behalf of the Working Group for Neonatal Screening in Paediatric Endocrinology of the European Society for Paediatric Endocrinology. Guidelines for a neonatal screening program for congenital hypothyroidism. Acta Paediatr 1999; 88(Suppl 432):13-4. 9. Working Group on Congenital Hypothyroidism of the Japanese Society for Pediatric Endocrinology and the Japanese Society for Mass Screening. Guideline for neonatal mass screening for congenital hypothyroidism. Clin Pediatr Endocrinol 1999; 8:51-5. 10. Selva KA, Mandel SH, Rien L, Sesser D, Miyabira R, Skeels M, et al. Initial treatment dose of L-thyroxine in congenital hypothyroidism. J Pediatr 2002;141: 786-92. 11. Rovet JF, Ehrlich R, Sorbara D. Intellectual outcome in children with fetal hypothyroidism. J Pediatr 1987;110:700-4. 12. Kooistra L, Laane C, Vulsma T, Schellekens JMH, van der Meere JJ, Kalverboer AF. Motor and cognitive development in children with congenital hypothyroidism: a long-term evaluation of the effects of neonatal treatment. J Pediatr 1994;124:903-9. 13. Germak JA, Foley TP. Longitudinal assessment of L-thyroxine therapy for congenital hypothyroidism. J Pediatr 1990;117:211-9. 14. Hanukoglu A, Perlman K, Shamis I, Brnjac L, Rovet J, Daneman D. Relationship of etiology to treatment in congenital hypothyroidism. J Clin Endocrinol Metab 2001;86:186-91. 15. Panoutsopoulos G, Mengreli C, Ilias I, Batsakis C, Christakopoulou I. Scintigraphic evaluation of primary congenital hypothyroidism: results of the Greek screening program. Eur J Nucl Med 2001;28:529-33. 16. Schoen EJ, Clapp W, To TT, Fireman BH. The key role of newborn thyroid scintigraphy with isotopic iodide (123I) in defining and managing congenital hypothyroidism. Pediatrics 2004;114:e683-8. 17. Kojima K, Ogawa E, Katsushima Y, Fujiwara I, Ohura T, Iinuma K. Ultrasonographic findings in neonates screened for congenital hypothyroidism. Clin Pediatr Endocrinol 2002;11:93-7. 18. Matsuura N, Harada S, Ohyama Y, Shibayama K, Fukushi M, Ishikawa N, et al. The mechanisms of transient hypothyroxinemia in infants born to mothers with Graves’ disease. Pediatr Res 1997;42:214-8. 19. Kreisner E, Camargo-Neto E, Maia CR, Gross JL. Accuracy of ultrasonography to establish the diagnosis and aetiology of permanent primary congenital hypothyroidism. Clin Endocrinol (Oxf) 2003;59:361-5. 20. Ohnishi H, Sato H, Noda H, Inomata H, Sasaki N. Color Doppler ultrasonography: diagnosis of ectopic thyroid gland in patients with congenital hypothyroidism caused by thyroid dysgenesis. J Clin Endocrinol Metab 2003;88:5145-9.
The Journal of Pediatrics • July 2008
AND IKUMA
FUJIWARA, MD
Objective To identify those infants who need a higher starting dose of levothyroxine (l-T4) for early normalization of thyroid-stimulating hormone (TSH) level. Study design TSH levels at 2 time points (1 to 3 weeks and 3 to 5 weeks) after l-T4 therapy at a starting dose of 8 to12 g/kg/day were evaluated retrospectively in 22 hypothyroid infants screened for congenital hypothyroidism (CH) in terms of etiology as determined by ultrasonography (US), the size of distal femoral epiphysis (DFE), and initial thyroid function. Results The infants with a noneutopic thyroid or small DFE exhibited significantly higher posttherapeutic TSH levels compared with the other infants. Eight of the 9 infants who failed to achieve normalized TSH values at 1 to 3 weeks had noneutopic thyroid. All of the infants with eutopic thyroid exhibited normalized TSH at 3 to 5 weeks, and a significantly greater proportion of the infants with eutopic thyroid exhibited normalized TSH at 1 to 3 weeks compared with those with noneutopic thyroid. Stepwise regression analysis demonstrated that US etiology was a significant independent variable for normalization of TSH at 1 to 3 weeks. Conclusions US examination to identify eutopic or noneutopic thyroid provides useful information for determining the starting dose of l-T4 in hypothyroid infants. (J Pediatr 2008;153:101-4) lthough the introduction of newborn screening has greatly improved the intellectual outcome in congenital hypothyroidism (CH), closing the remaining subtle neurologic gap in children with severe CH remains a work in progress.1,2 Using a higher starting dose to more quickly normalize thyroid-stimulating hormone (TSH) levels could result in normalization of developmental IQ even in those with severe CH.3-6 Consequently, an initial levothyroxine (l-T4) dose of 10 to 15 g/kg/day is now recommended.7-9 However, it is not clear whether replacement therapy should start with 10 or 15 g/kg/day; the recommendation is vague, specifying only “depending on the severity of the initial hypothyroidism”7 or “in the severe cases.”9 Furthermore, although it has been shown that an initial l-T4 dose of 12 to 17 g/kg/day can raise the TSH level to the target range in 2 weeks,10 all hypothyroid infants might not need the full dose. In CH, disease severity is indicated by etiology and skeletal maturity, in addition to the degree of initial hyperthyroidism. Because infants with athyreosis exhibit delayed skeletal maturity11 and lower initial free thyroxine (fT4) levels12 and take longer to achieve normal TSH values13,14 with more frequent dosage adjustment14 compared with infants with ectopic thyroid or dyshormonogenesis, having a different management strategy for each etiology is clearly the ideal situation. To obtain a precise diagnosis, 123I-scintigraphy (but not technetium scanning) has been recommended.7,15,16 However in Japan, as a practical matter, performing 123I-scintigraphy during the early neonatal period is difficult without delaying the start of treatment; thus, it is often performed later (eg, at age 5 years). Thyroid ultrasonography (US), an alternative approach, is not considered to give additional information useful for management, probably due to its weakness in detecting the ectopic gland, the most common cause of CH. Here we report the different responses to initial treatment in infants with CH depending on their US-determined etiology.
A
METHODS Screening Procedure In Japan, neonatal CH screening is based primarily on the measurement of TSH. Any infant with a TSH value ⱖ30 mU/L in a dried blood sample on filter paper at age
CH DFE fT4
Congenital hypothyroidism Distal femoral epiphysis Free thyroxine
l-T4 TSH US
Levothyroxine Thyroid-stimulating hormone Ultrasonography
From the Department of Pediatrics, Tohoku University School of Medicine, Sendai, Japan. Submitted for publication May 23, 2007; last revision received Oct 31, 2007; accepted Dec 5, 2007. Reprint requests: Eishin Ogawa, 2-1 SeiryoMachi, Aoba-ku, Sendai 980-8575, Japan. E-mail: [email protected]. 0022-3476/$ - see front matter Copyright © 2008 Mosby Inc. All rights reserved. 10.1016/j.jpeds.2007.12.015
101
5 days is immediately referred to a pediatrician for diagnostic evaluation. Any infant with a borderline result (10 ⱕ TSH ⬍ 30 mU/L) undergoes a second heel puncture and is referred if the result is again ⱖ10 mU/L.
Subjects A total of 195,621 neonates were screened for CH in Miyagi Prefecture, Japan, during the 9-year period between April 1995 and March 2004. The referral rate during this period was 0.22%. A total of 386 neonates (90% of those with positive results) were referred to Tohoku University Hospital. Of these, 130 infants were treated with l-T4 either transiently or permanently: 59 with mild CH (pretreatment serum TSH ⬍20 mU/L), 40 with moderate CH (pretreatment serum TSH 20 to 80 mU/L), and 31 with severe CH (pretreatment TSH ⬎80 mU/L). Only those with severe CH were candidates for this study, because the study’s major objective was to identify those infants who should require a higher dose (eg, 15 g/kg/day). Those infants referred after the second heel puncture were excluded, because their hypothyroidism could be affected by environmental factors after birth (eg, iodine exposure). Thus, the enrollment criteria were as follows: (1) abnormal initial screening result (dried blood TSH ⱖ30 mU/L); (2) confirmatory serum TSH ⬎80 mU/L and fT4 ⬍1.2 ng/dL; and (3) an l-T4 starting dose of ⱖ8 g/kg/day. A total of 67 infants exhibited abnormal initial results; of these, 25 were selected based on criterion 2, and 3 were excluded based on criterion 3. One of these latter infants could not tolerate oral l-T4 due to gastrointestinal malformation, and the other 2 were treated with lower doses because of suspected transient CH born to mothers with Graves’ disease taking antithyroid medication. Thus, a total of 22 infants were evaluated. These infants were allocated according to factors indicating disease severity, such as confirmatory fT4 and TSH levels, size of the distal femoral epiphysis (DFE), and US-derived etiology, as described later. Ultrasound Scanning A real-time mechanical sector scanner (EUB-8000 ultrasound scanner; Hitachi, Tokyo, Japan) with a digital beamformer system and a 10-MHz probe was used for the thyroid scan. A thyroid gland clearly visible at the normal site on US was classified as eutopic. In our previous 2-year survey in 97 infants with positive screening results, 88 had eutopic thyroid (5 as enlarged thyroid), 7 had noneutopic thyroid, and 2 had hemiagenesis.17 Eutopic thyroid suggests normal thyroid or dyshormonogenesis as a cause of CH, and noneutopic thyroid suggests dysgenesis, such as occurs in athyreotic, hypoplastic, or ectopic thyroid. Of the 22 infants in the present study, 8 had eutopic thyroid and 14 had noneutopic thyroid. A total of 16 infants with noneutopic thyroid were identified (about 1 in 11,000) during the 9-year study period; 15 of these demonstrated initial abnormal screening results, 14 of whom were enrolled in the present study. 102
Ogawa, Kojima-Ishii, and Fujiwara
Initial Treatment and Management Replacement therapy was started promptly (on day 15.5 ⫾ 6.7 on average) with an initial dose of 8 to 12 g/kg/day (mean, 9.66 ⫾ 0.98 g/kg/day) after blood sampling, X-ray to evaluate skeletal maturity, and thyroid US. TSH and fT4 measurements were obtained between 1 and 3 weeks and again between 3 and 5 weeks after the treatment. The l-T4 dose was increased in 6 infants but decreased in none, based on the TSH and fT4 values at 1 to 3 weeks. The exact values of TSH and fT4, along with the numbers of infants whose TSH levels reached within the target range, were compared among the groups. All infants with mild or moderate CH (serum TSH ⬍80 mU/L), who were excluded from this evaluation, were treated with ⱕ10 g/kg/day, and all reached the target TSH range at 3 to 5 weeks. Skeletal Maturity Skeletal maturity was assessed based on the size of the DFE, defined as the ratio of the transverse diameter between the distal epiphysis and the metaphysis of the femoral bones. The mean DFE size in term Japanese newborns is 0.31 ⫾ 0.07118; a value ⬍⫺2 standard deviations (0.168) indicates a small DFE. A small DFE was found in 6 of the 14 infants with noneutopic thyroid and in none of the 8 infants with eutopic thyroid. Hormone Measurements Serum TSH and fT4 levels were measured by sandwich enzyme imuumoassay (AxSYM; Dainabot, Tokyo, Japan). The normal values for infants in our hospital are as follows: TSH, 0.3 to 10.0 mU/L at age 1 month and 0.3 to 7.5 mU/L at age 2 months; fT4, 1.2 to 2.0 ng/dL for the entire period. Serum fT4 levels at confirmation ⬍0.5 ng/dL were designated “very low” (n ⫽ 12), and those ⱖ0.5 ng/dL (but ⬍1.2 ng/dL) were designated “low.” Very low fT4 was found in 5 of the 8 infants with eutopic thyroid and in 7 of the 14 infants with noneutopic thyroid. Serum TSH levels at confirmation ⬎300 mU/L were designated “very high” (n ⫽ 12), and those ⱕ300 mU/L (but ⬎80 mIU/L) were designated “high” (n ⫽ 10). Very high TSH was found in 2 of the 8 infants with eutopic thyroid and in 8 of the 14 infants with noneutopic thyroid. The target range for TSH is ⬍10 mU/L at 1 to 3 weeks and ⬍7.5 mU/L at 3 to 5 weeks. Statistical Analysis The differences in serum levels of TSH and fT4 at 3 time points in the various categories of patients were analyzed using the nonparametric Mann-Whitney U test. The distribution of infants achieving the target TSH range at 2 time points after l-T4 therapy was compared among the groups using 2 tests. Stepwise regression analysis was performed to identify the variables contributing to the definition of normalized TSH. Statistical significance was accepted at P ⬍ .05. The Journal of Pediatrics • July 2008
Figure 1. TSH levels (mU/L) at confirmation (A), 1 to 3 weeks (B), and 3 to 5 weeks (C). The data are medians, 25th and 75th percentiles, and ranges. Asterisks denote statistical significance (*P ⬍ .05; **P ⬍ .01; ***P ⬍ .001). The numbers in parentheses are the patient numbers.
RESULTS The serum TSH levels for each patient group are shown in Figure 1. Concentrations of serum confirmation TSH and fT4 levels differed significantly between the groups classified in terms of TSH and fT4 values, but not between the groups classified according to US-derived etiology and DFE size. The postreatment serum fT4 levels did not differ between the groups in each category (data not shown). The posttreatment serum TSH levels differed significantly between the groups categorized based on US-derived etiology and DFE size. TSH levels at 1 to 3 weeks and 3 to 5 weeks were higher in the noneutopic and small DFE groups compared with the eutopic and normal DFE groups. Thirteen infants (59.1%) reached the target TSH range at 1 to 3 weeks, and 17 (77.3%) did so at 3 to 5 weeks. The rates of TSH normalization for the groups of each category are shown in Figure 2. According to the US-derived etiology (but not the other categories), the distribution of infants with normalized TSH differed significantly at 1 to 3 weeks (2 ⫽ 4.197; P ⬍ .05), reflecting the greater proportion of infants with eutopic thyroid reaching the target range of TSH. In the eutopic group, 7 of 8 infants achieved normalized TSH at 1 to 3 weeks; the remaining infant, who had an abnormal value at 1 week, reached the target range at 4 weeks posttreatment without a dosage increment. In contrast, more than 50% of the infants with noneutopic thyroid failed to normalize TSH at 1 to 3 weeks, and 5 of these infants exhibited abnormal values at 3 to 5 weeks despite a dosage increment. A similar trend was observed in the other categories, with those infants with a small DFE, very low fT4, or very high TSH exhibiting lower rate of TSH normalization; none of these differences were statistically significant, however. Because those infants with noneutopic thyroid tended to have a small DFE (0.171 ⫾ 0.011 vs 0.241 ⫾ 0.005; P ⫽ Ultrasound Appearance of Thyroid Tissue in Hypothyroid Infants
Figure 2. Proportion of the infants who reached the target range of TSH levels (normalization rate) at 1 to 3 weeks (A) and 3 to 5 weeks (B). The white bars indicate children with normalized TSH; the solid bars indicate those with normalization failure of TSH. An asterisk (*) denotes statistical significance (P ⬍ .05). The numbers in parentheses are the patient numbers.
.1150) and more severe initial CH (baseline TSH higher in the noneutopic group, albeit not significantly so; see Figure 1), these categories are associated with one another. This finding led us to perform stepwise regression analysis to identify the significant variables defining normalization of TSH. Only the US-derived etiology was found to be a significant independent variable for normalization of TSH at 1 to 3 weeks (P ⬍ .05) and a nearly significant independent variable at 3 to 5 weeks (P ⫽ .058).
DISCUSSION In the present study, l-T4 therapy at a mean dose of 9.66 g/kg/day failed to normalize TSH levels at 3 to 5 weeks posttreatment in approximately 25% of the infants. Because the current goal of initial treatment of CH is to normalize TSH by age 1 month,7,8 in an effort to achieve good intellectual outcome even in the most severely affected infants, the higher dose was indicated for those who did not respond to the initial therapy, and characterization of the nonresponders was the study’s major aim. Responses to treatment differ dependent on the etiology, with infants with athyreosis requiring longer to achieve normalized TSH than those with ectopic thyroid and dyshormonogenesis.13,14 Therefore, we expected that the nonresponders would be demonstrated to have noneutopic thyroid by US. Indeed, 8 of the 9 nonresponders at 1 to 3 weeks and all of the nonresponders at 3 to 5 weeks had noneutopic thyroid. In addition to the fact that those infants with eutopic thyroid exhibited prompt normalization of TSH (Figures 1 and 2), and that stepwise regression analysis demonstrated that US-derived etiology is a significant factor in normalizing TSH, US also helped identify those infants requiring the higher l-T4 dosage. Although determining the etiology by 123I-scintigraphy is certainly ideal, US is an alternative technique when 103
123
I-scintigraphy is not available. US can accurately detect the thyroid gland in its normal location, but may fail to detect an ectopic gland,19 although a recent study has demonstrated markedly improved sensitivity of color Doppler US in diagnosing ectopic thyroid tissue.20 This weakness is probably the major reason why US has not been emphasized to date. Despite this drawback, however, the present study has demonstrated that a rough estimation with US, whether or not the thyroid is located normally, can provide useful information when determining the starting dose of l-T4. Furthermore, because it can be assumed that a small, ectopic thyroid with severe initial hyperthyroidism will clinically resemble athyreosis in terms of disease severity, the ability to differentiate a small ectopic thyroid from athyreosis should not be necessary for initial management planning. We propose providing a rough estimation with US before starting therapy, followed by subsequent precise diagnosis with 123I-scintigraphy, as is practical. DFE size, a measurement of skeletal maturity at birth, is considered to reflect the severity of CH in utero, with a caveat that because DFE depends on gestational age, a small DFE may be found in infants of shortened gestational age. The greater proportion of infants with athyreosis exhibit delayed skeletal maturation, and those with delayed skeletal maturation have intellectual impairment as well.11 Thus, DFE size is a factor in determining disease severity; however in our subjects, only those with noneutopic thyroid exhibited a small DFE. Although none of our infants with eutopic thyroid had a small DFE despite severe initial hyperthyroidism, infants with total dyshormonogenesis may exhibit a small DFE, and some infants with eutopic thyroid may need the maximum initial dose. A combination of US-derived etiology and DFE size might be advantageous for evaluating a nonresponder; such an evaluation was not feasible in our study due to the small number of subjects. In conclusion, although our results are preliminary (because of the small number of subjects and retrospective manner of the study design), they indicate that US to identify eutopic or noneutopic thyroid can be useful in determining the optimum starting dose of l-T4 for CH. When the USbased diagnosis is eutopic thyroid, an initial dose of approximately 10 g/kg/day appears to be sufficient (except possibly in those infants with a small DFE). In contrast, when the US-based diagnosis is noneutopic thyroid, an initial l-T4 dose
104
Ogawa, Kojima-Ishii, and Fujiwara
of up to 15 g/kg/day is advisable, to quickly normalize TSH levels.
REFERENCES 1. Hindmarsh PC. Optimisation of thyroxine dose in congenital hypothyroidism. Arch Dis Child 2002;86:73-5. 2. Rovet JF. In search of the optimal therapy for congenital hypothyroidism. J Pediatr 2004;144:698-700. 3. Dubuis JM, Glorieux J, Richer F, Deal CL, Dussault JH, van Vliet G. Outcome of severe congenital hypothyroidism: closing the developmental gap with early high-dose levothyroxine treatment. J Clin Endocrinol Metab 1996;81:222-7. 4. Bongers-Schokking JJ, Koot HM, Wiersma D, Verkerk PH, de Muinck KeizerSchrama SM. Influence of timing and dose of thyroid hormone replacement on development in infants with congenital hypothyroidism. J Pediatr 2000;136:292-7. 5. Selva KA, Harper A, Downs A, Blasco PA, LaFranchi SH. Neurodevelopmental outcomes in congenital hypothyroidism: comparison of initial T4 dose and time to reach target T4 and TSH. J Pediatr 2005;147:775-80. 6. Simoneau-Roy J, Marti S, Deal C, Huot C, Robaey P, Van Vliet G. Cognition and behavior at school entry in children with congenital hypothyroidism treated early with high-dose levothyroxine. J Pediatr 2004;144:747-52. 7. American Academy of Pediatrics. Update of newborn screening and therapy for congenital hypothyroidism. Pediatrics 2006;117:2290-303. 8. Toublanc JE on behalf of the Working Group for Neonatal Screening in Paediatric Endocrinology of the European Society for Paediatric Endocrinology. Guidelines for a neonatal screening program for congenital hypothyroidism. Acta Paediatr 1999; 88(Suppl 432):13-4. 9. Working Group on Congenital Hypothyroidism of the Japanese Society for Pediatric Endocrinology and the Japanese Society for Mass Screening. Guideline for neonatal mass screening for congenital hypothyroidism. Clin Pediatr Endocrinol 1999; 8:51-5. 10. Selva KA, Mandel SH, Rien L, Sesser D, Miyabira R, Skeels M, et al. Initial treatment dose of L-thyroxine in congenital hypothyroidism. J Pediatr 2002;141: 786-92. 11. Rovet JF, Ehrlich R, Sorbara D. Intellectual outcome in children with fetal hypothyroidism. J Pediatr 1987;110:700-4. 12. Kooistra L, Laane C, Vulsma T, Schellekens JMH, van der Meere JJ, Kalverboer AF. Motor and cognitive development in children with congenital hypothyroidism: a long-term evaluation of the effects of neonatal treatment. J Pediatr 1994;124:903-9. 13. Germak JA, Foley TP. Longitudinal assessment of L-thyroxine therapy for congenital hypothyroidism. J Pediatr 1990;117:211-9. 14. Hanukoglu A, Perlman K, Shamis I, Brnjac L, Rovet J, Daneman D. Relationship of etiology to treatment in congenital hypothyroidism. J Clin Endocrinol Metab 2001;86:186-91. 15. Panoutsopoulos G, Mengreli C, Ilias I, Batsakis C, Christakopoulou I. Scintigraphic evaluation of primary congenital hypothyroidism: results of the Greek screening program. Eur J Nucl Med 2001;28:529-33. 16. Schoen EJ, Clapp W, To TT, Fireman BH. The key role of newborn thyroid scintigraphy with isotopic iodide (123I) in defining and managing congenital hypothyroidism. Pediatrics 2004;114:e683-8. 17. Kojima K, Ogawa E, Katsushima Y, Fujiwara I, Ohura T, Iinuma K. Ultrasonographic findings in neonates screened for congenital hypothyroidism. Clin Pediatr Endocrinol 2002;11:93-7. 18. Matsuura N, Harada S, Ohyama Y, Shibayama K, Fukushi M, Ishikawa N, et al. The mechanisms of transient hypothyroxinemia in infants born to mothers with Graves’ disease. Pediatr Res 1997;42:214-8. 19. Kreisner E, Camargo-Neto E, Maia CR, Gross JL. Accuracy of ultrasonography to establish the diagnosis and aetiology of permanent primary congenital hypothyroidism. Clin Endocrinol (Oxf) 2003;59:361-5. 20. Ohnishi H, Sato H, Noda H, Inomata H, Sasaki N. Color Doppler ultrasonography: diagnosis of ectopic thyroid gland in patients with congenital hypothyroidism caused by thyroid dysgenesis. J Clin Endocrinol Metab 2003;88:5145-9.
The Journal of Pediatrics • July 2008