Congenital Hypothyroidism: An Evolving Common Clinical Conundrum

Newborn screening for congenital hypothyroidism, initiated in 1975 (1), has led to the recognition and treatment of affected infants before the onset of clinical signs and symptoms of hypothyroidism. Widespread screening has in large part reduced, if not eliminated, the adverse neurological and developmental effects of congenital hypothyroidism. Compared with the prescreening period, the worldwide reported incidence has more than doubled and is now estimated to be between 1:2000 and 1:4000 newborns. Although there is general agreement that treatment should begin as early as possible and with daily oral doses of T4 between 10 and 15 μg/kg (2), there is a remarkable lack of uniformity in the diagnoses assigned and in the reporting and follow-up of infants identified and treated with hormone replacement therapy. A report by Mengreli et al. in the current issue of JCEM (3) using a primary TSH screening method (at 3–5 d of age) reports a significant increase in the incidence of congenital hypothyroidism diagnosed when the neonatal TSH threshold limit was decreased from 20 to 10 mU/liter. With this change, 28% more infants were placed on replacement thyroid hormone (3). Not unexpectedly, this also resulted in a 10-fold rise in the recall rate of screened infants. At 3 yr of age, 151 of the 200 children identified as having congenital hypothyroidism were reevaluated. What was most remarkable is that 88.7% of the 151 reevaluated children were deemed to have permanent hypothyroidism. Thirty-one (20%) of the reevaluated children were premature infants, with 23 of these children having permanent hypothyroidism. One third of the babies with permanent hypothyroidism had screening TSH levels between 10 and 20 mU/liter and would therefore have been missed had the cutoff level remained at 20 mU/liter, and half of the premature babies with persistent congenital hypothyroidism had screening TSH levels of less than 20 mU/liter. Structural abnormalities of the thyroid gland were detected in 20% of the infants with screening TSH levels of less than 20 mU/liter. The diagnosis of permanent hypothyroidism was based on a serum TSH concentration of greater than the upper limit of normal for the assay (5 mU/liter) after discontinuation of therapy for at least 3 wk, the presence of a sibling with an increased serum TSH, or an abnormal thyroid gland on either imaging or functional tests. Imaging tests consisted of ultrasound and radioisotope screening and perchlorate discharge. Of note, the initial screening was obtained at 3–5 d of age, which is somewhat later than the age at which screening tests are obtained in the United States. By applying the TSH threshold of 10 mU/liter, the prevalence of permanent congenital hypothyroidism was 1:1,749, whereas that of transient congenital hypothyroidism was 1:14,154. These findings are consistent with those of Corbetta et al. (4), who reported an increase in the incidence of congenital hypothyroidism from 1:2654 to 1:1446 cases with a reduction in TSH cutoff from greater than 20 mU/liter to greater than 10–12 mU/liter. The increase was largely, but not exclusively, in those with eutopic thyroid glands. When reevaluated at 3 yr of age, 78% of these cases were thought to have permanent hypothyroidism. TSH levels greater than 10 mU/liter with low or normal free T4 were seen in 34% of the cases, and mild permanent thyroid dysfunction with TSH levels between 5 and 9.9 mU/liter with normal free T4 was seen in 44% of the cases. In the United Kingdom, the initial screening threshold adopted by the various neonatal screening services ranged from 5–12 mU/liter in samples collected between 5 and 8 d of life. All programs use similar laboratory technology to assay TSH. Korada et al. (5) suggested that a cutoff value of 10 mU/liter compared with 6 mU/liter may miss several babies with congenital hypothyroidism, although the number of false-positive term infants increased by 126%. In the United States, there has been a similar increase in the reported incidence (birth prevalence) of congenital hypothyroidism, without alteration in screening threshold levels. In New York State alone, using a primary T4 screen, the incidence of congenital hypothyroidism between 1978 and 2005 increased by 138% to approximately 1:1400. This figure is remarkably similar to that reported by Mengreli et al. (3) and Corbetta et al. (4) with the lower TSH cutoff levels. In the United States, with 58 million infants screened between 1987 and 2002, the incidence increased by 73% (6). Reports were recently published from a workshop held in 2008 at the Centers for Disease Control in Atlanta, Georgia, designed to explore the causes of the apparent increase in the prevalence of congenital hypothyroidism. Modifications and improvements in laboratory practices, population shifts, and the reporting of patients with transient hypothyroidism among those with congenital hypothyroidism were all thought to be insufficient explanations of the remarkable increase in prevalence (7–10). What became apparent was a lack of consistent definition for both congenital hypothyroidism and transient hypothyroidism. Similarly, there were no uniform protocols for testing and reporting of results among the states. I have previously argued for the inclusion of mild congenital hypothyroidism within the spectrum of congenital hypothyroidism (11). Because more and more sick as well as very premature infants are surviving, the number of diagnosed cases of thyroid dysfunction that may require treatment in the neonatal period is likely to increase substantially. Included may be patients with transient hyperthyrotropinemia and some with a late rise in TSH (infants who have an initial normal newborn screening but, after several weeks in the neonatal intensive care unit, develop an elevated level of TSH that may require intervention) (12). Some of the infants with mild hypothyroidism or with late rise of TSH may not have evidence of persistent hypothyroidism when reevaluated, usually after 3 yr of age. Genetic studies seem to validate the existence of mild or transient congenital hypothyroidism. Genetic defects in dual oxidase 2 gene and others have been associated with mild as well as transient congenital hypothyroidism (13–15). The article by Mengreli et al. (3) serves to highlight the need for consensus development regarding the diagnosis and treatment of congenital hypothyroidism. Agreement is crucial to define the conditions counted as congenital hypothyroidism (severe, permanent, mild, transient, late rise of TSH). Data are needed to provide evidence of whether or not screening should be designed to identify severe cases alone or all cases in need of treatment, even transiently. Uniformity is also needed regarding a protocol with which to evaluate the etiology of hypothyroidism, as well as the persistence of hypothyroidism. Genetic studies, visualization studies, including ultrasound and scanning procedures, and measurements of serum thyroglobulin levels may all contribute. TRH stimulation tests have been found to be useful in the evaluation of patients with heterozygous mutations of the TSH receptor gene resulting in hypothyroidism (16, 17). We had previously suggested TRH tests for infants with mild elevations of TSH without marked abnormal T4 and T3 values (18). TRH testing may also be useful for children being reevaluated after withdrawal of treatment to establish whether or not a slightly elevated TSH level is indeed indicative of hypothyroidism. TRH testing may have a crucial role in diagnosing central hypothyroidism and isolated TSH deficiency, a condition not identified by screening programs using a primary TSH screen (19). However, TRH is not currently available for use in the United States (20). It is apparent that congenital hypothyroidism has evolved into a condition, not unlike many others in clinical medicine, whose manifestations range from severe to mild. Advances in molecular diagnosis, including screening for transcription factor mutations and chromosomal aberrations by array comparative genomic hybridization, can certainly shed some light on a relatively few cases of congenital hypothyroidism (21, 22). However, the same genetic abnormality can result in a spectrum of clinical manifestations from severe to mild to unaffected. A defect in PAX 8 may result in a variety of phenotypes, ranging from a normal to an absent thyroid gland and from euthyroidism to severe hypothyroidism (23). Prospective, large-scale, long-term studies are needed to establish the benefits, if any, of the treatment of subtle cases of congenital hypothyroidism. Gathered evidence can then be used to pursue the personalized medical care that all infants deserve. Infants with severe as well as mild, permanent as well as transient hypothyroidism deserve standardized, quality, evidence-based clinical care. Whether or not it is the function of the newborn screening programs to identify and collect data on all of these patients, including those with central hypothyroidism, remains open to debate. I believe the convening of an international expert consensus-building conference to address the controversies surrounding congenital hypothyroidism is essential. I acknowledge the invaluable help of Tamar Simpser, without whose expertise this would not have been possible. Disclosure Summary: The author has nothing to disclose. For article see page 4283