Newborn screening for congenital hypothyroidism

Abstract

Introduction Congenital hypothyroidism (CH) is the commonest treatable cause of mental retardation. It is one of the most common disorders related to mental impairment and growth retardation in newborns. In many countries, neonatal thyroid screening programs are performed for early diagnosis and treatment of hypothyroidism. CH is usually sporadic and occurs in one in 3000-4000 infants. Most infants with CH are normal at birth and show no signs, emphasizing the importance of screening programs in early detection of CH (1,2). Newborn screening (NS) for CH is one of the major achievements of preventive medicine. Although since 1972 the problem of CH has been resolved in developed countries by the implementation of NS, the same cannot be said for developing countries that still have no NS programs for CH (2,3). Since diagnosis based on clinical findings is delayed in most instances because of few symptoms and signs, hypothyroidism in the newborn period is almost always overlooked, and delayed diagnosis leads to the most severe outcome of CH, namely, mental retardation. In a Danish study (4) conducted on infants born between 1970 and 1975, it was emphasized that only 10% of the affected infants were diagnosed within the first month of life, 35% within 3 months, and 70% within the first year. In the remainder of the infants, the diagnosis was delayed to the 3rd and 4th years of life. In a retrospective analysis of 1000 cases of CH from Turkey (5), the mean age at diagnosis was reported to be 49 months, and only 3.1% of cases were diagnosed within the first month, while 55.4% were diagnosed after 2 years of age. The first CH screening was performed by Dussault (6,7), in Quebec-Canada in 1972. They detected 7 hypothyroid infants among 47 000 newborns during a 3-year period. The high frequency of false positives delayed the diagnosis and increased the cost, and they therefore devised cut-offs values to be used for recall. Thyroid hormone and thyroid-stimulating hormone (TSH) levels were assessed in the recalled group of babies. In the meantime, radioactively labeled antibodies for determining thyroxine (T4) in dried blood spots were introduced regionally in the USA and in Europe. Screening programs for CH went parallel with screening programs of phenylketonuria. In the initial report by Dussault et al (8), the method was recommended as a confirmatory test, knowing that it would miss cases with hypothalamic-pituitary hypothyroidism, which they reported to constitute 10% of the cases. In 1976, Walfish (9) reported in the Lancet that cord blood TSH measurements had greater sensitivity and specificity as compared to cord blood T4 and spot blood (collected on 3 to 4 day old newborns) T4 results and that both false positives and costs were higher in the T4 method. This same author also suggested routine T4 supplemented by TSH estimation be used in mass screening. Although more sensitive, screening by T4 and TSH together is not cost-effective, therefore, mostly TSH, and rarely T4 screening, is used around the world. In Europe, TSH screening is preferred, whilst some centers in the USA prefer primary T4 testing supplemented by TSH (10,11,12,13,14). The recall rate for primary hypothyroidism in both approaches is 0.05%, and the rate of false positive results is higher using primary T4 strategy. TSH screening was shown to be more specific in the diagnosis of CH, while T4 screening was more sensitive in detecting newborns with rare hypothalamic-pituitary hypothyroidism but less specific with a high frequency of false positives mainly in low birth weight and premature babies. T4-binding globulin (TBG)-deficient babies who are euthyroid and who were not targets for NS, could also be detected by T4 screening. The Neonatal Thyroid Screening Conference held in Tokyo in 1982 recommended NS programs oriented to detect infants with elevated serum concentrations of TSH. It has been also suggested that this could be accomplished by measuring TSH in filter paper blood spot or by measuring T4 supplemented by TSH on the same blood spot of infants who have T4 values in the lower 3rd to 10th percentile (15).