Anastasia A. PANTAZAKI, Gregoris P. TSOLKAS and Dimitrios A. KYRIAKIDIS

LECTURE 2

SCREENING FOR CONGENITAL HYPOTHYROIDISM IN THE NEONATE

Congenital hypothyroidism (CH) is a disease long been known to result in mental retardation. It occurs in 1/3000 to 1/4000 newborn infants. Early diagnosis and treatment prevents brain damage and the ensuing mental retardation. Before the era of screening, only 10 % of the affected infants were diagnosed within the first month of life.

The method of screening for CH involves the determination of T₄ and/or TSH on dried blood spots collected on the second through fifth days of life. Ten-year analysis of our regional screening program for CH (since 1991) using primary TSH determination revealed a prevalence of 1/2512 for CH (1/3516 for dysgenesis, and 1/8791 for dyshormonogenesis). Prevalence of transient hypothyroidism was 1/1208. The recall rate was quite high at 2.8 % reflecting the iodine status of Turkey. Forty percent of infants with ectopy, and 20 % of those with dyshormonogenesis had initial serum T₄ levels within normal limits, in addition to 27.5 % of cases with transient hypothyroidism that needed hormone replacement in early life. Hence primary TSH screening is the preferred method for screening in Turkey.

Early detection of CH through newborn screening proved to be one of the great successes of preventive medicine. The screening should be oriented to detection of primary hypothyroidism. Expected standard for a screening test is a sensitivity approaching 100 %. However, owing to the element of human error and the potential for biological variations, no screening test could truly achieve 100 % over long time periods. Simultaneous measurement of both T₄ and TSH in dried blood spots is the most sensitive method to that effect, however it is not cost-effective. Alternatives are primary TSH, or primary T₄ supplemented by TSH.

İnci KARAARSLAN

Screening tests relied on the “one test-one disorder” concept until the introduction of tandem mass spectrometry into newborn screening in the 1990's. Profiling of amino acids and acylcarnitines in a single analysis has enabled newborn screening programs to expand testing to include up to 30 treatable inborn error of metabolism(IEM). Besides the increase in the number of diseases covered, tandem MS has also improved testing from an analytical point of view. İt is very specific and sensitive in its identification of the compounds. The false positive rates are lowered because disorders are identified not only on the basis of quantification of metabolites but also by the screening for a pattern of metabolite abnormalities as opposed to screening for a single metabolite and also by measuring metabolite ratios.

Between October 2001-August 2003, 12188 newborn (1-10 day old) were screened by tandem MS in our laboratory . %95.5 of the babies were healthy and had normal birth weight. % 4.5 of the babies either had birth weights less than 1500 gr , required neonatal intensive care, or had symptoms or familiy history of an IEM. Within the first group, three babies with PKU and one with Citrullinemia were identified. In the latter group, we identified 8 amino acid disorders, 4 urea cycle defects, 7 organic acidemias and 1 fatty acid oxidation defect.

Within the same period we also screened 1853 patients (age 11 days – 14 years old) who had clinical symptoms associated with IEM. We identified 15 amino acid disorders and 13 organic acidemias

The conclusions we can deduct from our experience with screening for IEM by tandem mass spectrometry are

1. The overall frequency of IEM is high in our country and newborn screening for these disorders at least in a selected high risk group will be cost effective both for the family and for the society in the long run.

2. Quite a number of treatable IEM can be rapidly diagnosed from a very simple sample , namely a dried blood spot which is both easy to obtain, to tranport and to store. This advantage should be made use of for screening IEM especially in states of emergency and in cases where laboratories capable of performing advanced metabolic tests are not readily available.

LECTURE 4

THE USE OF SPLIT-SAMPLE DESIGN FOR PERFORMANCE EVALUATION OF SCREENING KITS: A REAL LIFE STUDY AND EXCEL PROGRAMME FOR REFERENCE VALUE DETERMINATION OF nTSH MEASUREMENT

Prof. Dr. Tijen Tanyalçın (MD, PhD)

Ege University Medical School and Hospital Department of Biochemistry Bornova 35100 Izmir/Turkey

Laboratory medicine is an important discipline in health care with its remarkable effect on risk assessment, diagnosis of health and disease state and especially from newborn screening approach with its, retest, recall and follow-up procedures. This real life trial, emphasizes the need of split sample design evaluation of newly opened test kits. Quantitative measurement of phenylalanine and nTSH (neonatal thyroid stimulating hormone) were performed in both of the laboratories. After validation of calibrations were performed in the laboratory that used these industrially prepared screening kits for the first time, the same real newborn blood spot samples were analysed for phenylalanine and nTSH measurements in both of the laboratories and the obtained results were compared non parametrically and examined by the Deming regression graph and by the difference plot. There was no problem with the phenylalanine results, similar results were obtained for the same blood spot cards from both of the laboratories (P=0,496; bias estimation was 0,13). However, nTSH values were found to be significantly higher in the laboratory that used the nTSH kit for the first time. Although the validation of calibration of the nTSH kit was valid with its own control materials, split sample results showed that there was a significant difference between two laboratories (P=0,005; bias estimation was 28,6). This work implies that acceptable comparability of split sample design analysis is strictly needed for testing the analytical performance of the industrially prepared tests kits and this can be achieved only by certified reference materials.

Health-associated reference values are universally needed in clinical chemistry.

The aim of this study was to establish the reference intervals of two populations from data obtained by the mass screening of newborn babies and to demonstrate how to determine 95 % confidence intervals around the lower and upper limits of reference values from values that are not normally distributed. This experiment shows a way to define the rank numbers for n>1000 and to obtain reference values with 95% confidence intervals for lower and upper reference limits.