Revised Guidelines for Neonatal Screening Programmes for Primary Congenital Hypothyroidism

Since the first guidelines for neonatal screening for congenital hypothyroidism (CH) were issued by ESPE in 1993 [1], there have been considerable advances in our understanding of CH and our appreciation of the various geographical and logistic difficulties involved. Therefore, an updating of the guidelines is overdue. Experience from countries where screening began in the late 1970s and early 1980s has indicated that treatment should be started no later than the first 2 weeks of life using a ‘high’ dosage regime of L-thyroxine® (10–15 µg/kg/day). It has also been shown that the quality of long-term outcome is closely related to the quality of follow-up. In Eastern Europe, screening programmes for CH have either been started or will start soon in almost all countries, and although many programmes are operating satisfactorily, it is important to standardise screening procedures and management of suspected cases as much as possible in order to optimise outcome. A degree of uniformity throughout Europe would not only facilitate early detection and treatment of individual patients but give insight into the economic and epidemiological aspects of the screening programmes as well as the epidemiological aspect of CH.Adequate management of neonatal CH requires efficient capillary blood sampling, adequate testing, prompt communication of the screening results to a reference centre, accurate diagnosis and advice on treatment and follow-up by a paediatrician with endocrine expertise.The neonatal screening programme should be combined as much as possible with screening for other disorders, such as metabolic diseases and congenital adrenal hyperplasia, in order to minimise the discomfort of multiple blood samples for the babies, avoid multiple mailing and to reduce costs and risks of administrative errors.In Europe, two screening methods, depending on the aim of the screening, are used. If the aim of screening for CH lies in the prevention of severe mental retardation caused by primary CH, then thyroid-stimulating hormone (TSH) measurement alone will be used as the primary screening parameter. The TSH screening method is the most frequently used, because it is cheaper and easier to perform than the T4 method. Despite the elevated TSH levels seen just after birth in normal neonates, the TSH assay is sensitive enough to distinguish between normal newborns and neonates with primary CH to such an extent that the need for repeated measurements is low (recall rate 0.05% when TSH cut-off 15 mU/l of whole blood is used if measured with immunofluorometric methods, and 20 mU/l if measured radioimmunologically). However, the TSH method does not detect hypothalamo-pituitary forms of CH, and it has to be mentioned that some patients with central hypothyroidism will suffer from severe CH, but due to the low frequency of this disorder, it does not seem not to be justified to include T4 determinations into a TSH screening programme.The sensitivity of the TSH method for primary CH is 97.5%, the specificity is 99%. If the aim of the programme also comprises the detection of other forms of CH, the following method may be used: The T4 method is combined with either TSH determination or, in some centres, with both TSH and thyroid binding globulin (TBG) measurement. T4 is measured in all blood samples, TSH in the 20% lowest T4 samples (≤–0.8 SD) and TBG in the 5% lowest T4 values (≤–1.6 SD). This method of screening results in a sensitivity of 98% and a specificity of over 99% for all forms of CH [2]. The recall rate for this method is 0.4%, and the percentage of patients directly referred to a paediatrician 0.2%. Both methods are performed on dried capillary blood spots from heel puncture. Duplicate measurements are strongly advised.The screening results from both methods can be used to monitor iodine supply in the newborn population. This is an important issue in the many European countries which are still iodine deficient.It is usually recommended that capillary blood from a heel stab is performed before or at latest by day 5 (taking day of birth as day 1) in order to avoid unacceptable delay in diagnosis and treatment. If CH screening is combined with screening for congenital adrenal hyperplasia, capillary sampling should be performed not later than day 3. If, in the case of early discharge, sampling is performed 24 h after birth, this does not increase the recall rate provided that a sufficiently sensitive TSH method is used. The results of T4-TSH-TBG screening are not adversely affected either by early discharge. Indeed, cord blood sampling is possible for both methods, but this is too early for other neonatal screening programmes, i.e. phenylketonuria or congenital adrenal hyperplasia.Pre-term babies, especially those born before 32 weeks gestation, are particularly susceptible to transient hypothyroidism, nad also often show low TSH and T4 concentrations during the first weeks after birth, sometimes followed by TSH elevation. Therefore, the screening test should be repeated at 32 weeks gestation in these babies. Also, after the use of dopamine [3]in pre-term babies or in very sick term neonates, it is advisable to repeate the screening test twice after 7 or 14 days. In programmes where the T4-TSH-TBG method is used, repeated testing requires TSH measurement only. In cases where a blood transfusion or exchange transfusion is to be performed, venous blood for screening should be drawn before the intervention.Iodine contamination is the most fequent cause of neonatal transient hypothyroidism in term neonates. Therefore all newborns receiving significant amounts of iodine-containing compounds (e.g. disinfectants, contrast agents, amiodarone) should have their thyroid status closely monitored. In cases where the mother has received iodine, amiodarone or other anti-thyroid drugs, the neonate may have a positive screening test, and in some cases, the transient hypothyroidism can be so marked that treatment is indicated for a period of time.In several nurseries, capillary blood sampling using heel pricks has been switched for venous sampling (dorsal side of hand). This way is less harmful and more efficient than heel prick. Furthermore, anaesthetic ointments are being licensed for neonates which are efficient on venous punctures on the hand but less on the heel prick. In the absence of systematic and compared studies on the influence of both methods on results, no precise recommendations should be drawn for the present time. But as in phenylketonuria screening, small but significant alterations are observed depending on the method of sampling; this point should be taken in consideration by the laboratories in which the samples are obtained by both methods depending on the origin of the nurseries.Centralised screening laboratories should cover large enough regions, the optimal coverage being 100,000 newborns per year. If laboratories handle numbers below about 50,000 newborns per year, this may lead to inefficient screening due to a poor cost benefit ratio and to insufficient basis for quality control and statistical analysis. Screening laboratories based on less than 50,000 newborns per year may have to be accepted in countries with poorly developed postal or alternative communication systems or big countries with scarce widespread population. Otherwise, in many European countries, a single centralised screening laboratory is preferable.A system of quality control is an essential component of every screening programme. Quality control includes participation of screening laboratories in national and European control programmes to meet the requirements of sample processing, training of personnel and audit of performance. The results of screening and quality control should be reported yearly to the national screening authorities. It is strongly recommended that all pathological results and all epidemiologically interesting cases are reported to a central database which is handled by the central screening laboratory. Feedback information on screened cases sent by the paediatricians is an essential point for quality control.It is the responsibility of the screening laboratory to perform the thyroid hormone measurements on the screening blood sample as quickly as possible and for the results to be promptly reported to a reference centre. This centre is responsible for the further management of suspected cases, including recall for a blood sample or direct referral to a paediatrician if there is a strong suspicion of CH. Confirmation of the diagnosis by measurement of thyroid hormone concentration and TSH in venous blood can be done either by the screening laboratory or by the neonate’s local hospital. In both cases, the local paediatrician who is in charge of the baby should have access to a paediatrician with expertise in paediatric endocrinology, in order to ensure an appropriate standard of clinical evaluation, diagnosis, treatment and follow-up.Any value above the cut-off for TSH or below the cut-off for T4 should be confirmed by measurement in a venous sample in which at least TSH, T4 and free T4 analysis need to be measured. It is also recommended that blood be taken for thyroglobulin, TBG and thyroid autoantibodies. Evaluation of the aetiology and severity of the hypothyroidism can also be obtained by ultrasound examination of the thyroid gland, and by assessment of skeletal age with an X-ray of the knee. When severe CH is suspected, the results of the confirmation test should be available within a few hours, and in all cases within 24 h. If this cannot be achieved. Treatment can be initiated when adequate blood samples have been collected while the analytical results are pending. When the diagnosis of CH is confirmed, treatment should be started without delay. Further aetiologic investigations should not be allowed to postpone treatment, and most can be performed during the first week of therapy when TSH is still elevated, or at a later date, e.g. after the second or third year when therapy can be stopped safely for some weeks without damage to the central nervous system. In cases where CH is not confirmed, i.e. when a very high capillary TSH is followed by a normal venous TSH, there is a possibility that two samples are exchanged. To investigate this possibility, all samples in the same series should be checked again to detect for a falsely negative case, i.e. an infant with CH but normal screening values.In case of particular situations: e.g. consanguinity or occurrence of several cases in the same kindred and association with somatic malformations; further investigations (DNA) for specific gene mutations in specialised laboratories are recommended.The local paediatrician, together with the consulting paediatric endocrinologist, should make the decision as to when to treat, i.e. as soon as the diagnosis is either proven or extremely likely (e.g. screening TSH ≥50 mU/l and/or T4 <25 nmol/l, no thyroid gland visible on the ultrasound examination, skeletal age delayed by more than 6 weeks, clinical signs of CH). It is very important for patients with severe hypothyroidism to be treated no later than the first 2 weeks of life. The treatment of choice is L-thyroxine®. Treatment with T3 is obsolete. The starting dose of L-thyroxine is 10–15 µg/kg/day, to be given in 1 dose per day (in France a mean dose of L-thyroxine in solution: 8 µg/kg/day is sufficient to normalise TSH within 2 weeks).Treatment is monitored by regular measurement of plasma (free) T4 and TSH concentrations. It is advisable to measure these parameters weekly or fortnightly after initiation of therapy to confirm an adequate rise in free T4 concentration and normalisation of TSH. Optimisation of free T4 (>20 pmol/l) can be achieved within the first 2 weeks of treatment, and satisfactory reduction in TSH (<10 mU/l) within the first month. It has been shown that infants with plasma T4 concentrations during the first year below 130 nmol/l and/or elevated plasma TSH levels (>10 mU/l) have lower IQ scores than patients with normal concentrations. Therefore, (free) T4 concentrations should be kept in the upper normal range and TSH values between 0.5 and 5 mU/l.When investigation of the aetiologic diagnosis was not possible or was not completed in the neonatal period, or when there is doubt about the diagnosis, re-evaluation can be planned after the age of 2 or 3 years. This is done by a withdrawal of L-thyroxine for a 4-week period. Before and after the withdrawal, (free) T4 and TSH concentrations are measured. If the concentrations are normal and unchanged and remain normal during the following months, CH in the neonatal period was transient. If (free) T4 concentrations have decreased and TSH concentrations have risen, CH is proved and the planned investigations (ultrasound, scintigraphy, 24-hour urine collection) can be performed.Periodic and systematic mental development should be checked by a trained psychologist, and results obtained should be compared to a control group. As the prevalence of hearing loss is about tenfold higher in infants with CH, the detection of troubles by otoacoustic emissions is highly recommended during the first 2 months of life or not later than at the age of 3 months in order to avoid difficulties related to impaired speech development in late-diagnosed cases which was the most usual in the absence of screening [4, 5].