Abstract
Objectives
Clinical laboratories store filter paper samples used in neonatal screening for various periods of time after performing hormonal measurements. However, due to lack of data concerning specimen stability, it is unclear for how long these samples should be stored. The objective of this study is to determine the stability and reproducibility of thyroid-stimulating hormone (TSH), thyroxine (T4) and 17-hydroxyprogesterone (17-OHP) measurements in filter paper blood samples stored for up to 60 months.
Methods
Two hundred and twenty-eight blood samples, drawn between the second and the fourth day of life, were divided into seven distinct groups and kept at 4–8°C for one day or 2, 12, 24, 36, 48 or 60 months after basal hormonal measurements. In each group, TSH, T4 and 17-OHP levels were initially assayed 24–48 hours after collection (basal) and repeated once at the end of storage timing. All the measurements were performed by time-resolved fluorometry (1235 AutoDELFIA, Wallac Oy, Turku, Finland). Repeated and basal levels of each hormone were compared within the same group by Student's paired t-test. Differences were considered significant at P < 0.05.
Results
Compared with basal measurements, TSH and T4 levels declined significantly only when these hormones were re-assayed at 48 or 60 months of sample storage. In contrast, 17-OHP concentrations decreased earlier, starting at 24 months and continuing throughout the remaining period.
Conclusion
Our data suggest that neonatal screening of filter paper samples kept at 4–8°C are reliable for repeating the hormonal measurements when specimens are stored for up to one year, in the case of 17-OHP, or three years, in the case of T4 and TSH.
INTRODUCTION
Neonatal screening for congenital hypothyroidism is legally required in most countries due to the high prevalence of this disorder (1:4000). 1,2 This is usually done by measuring thyroid-stimulating hormone (TSH) and/or thyroxine (T4) in blood samples collected in filter papers drawn in the nursery, ideally between the second and fourth day of life. 3 Recently, some neonatal programmes have introduced the concomitant measurement of 17-hydroxyprogesterone (17-OHP) to screen for congenital adrenal hyperplasia (CAH) due to 21-hydroxylase deficiency, a virilizing and salt wasting disorder with an incidence of 1:5000 to 1:20,000. 4–6
Clinical laboratories store filter paper samples for long periods of time after performing neonatal measurements because of the potential need to repeat them in the future. 7 However, due to lack of data, it is unclear for how long neonatal samples could be stored and still maintain hormonal stability. To address this question, we compared the results of repeated measurements of TSH, T4 and 17-OHP in blood samples drawn in a filter paper and kept at 4–8°C for up to five years as part of the neonatal screening programme in an academic hospital (Hospital das Clinicas da Faculdade de Medicina da Universidade de Sao Paulo, Brazil).
PATIENTS/METHODS
Two hundred and twenty-eight samples were drawn between the second and fourth day of life by heel-prick with an automated lancet device (DBS Puncher, Wallac Oy, Turku, Finland) and spotted in 12.7 mm preprinted circles on SS903 filter paper. All the specimens were obtained from newborns that were appropriate for gestational age and kept at 4–8°C in sealed plastic bags, which were packed in closed boxes without added desiccants. The temperature of the refrigerators was monitored daily and varied from 2 to 8°C.
Blood spotted in filter paper was limited to a small circle, enabling only a few hormonal measurements per sample. To circumvent this problem, samples selected for this study were divided into seven groups of 27–41 different specimens. In each group, samples underwent basal measurements of TSH, T4 and 17-OHP, which were performed 24–48 hours after collection. They were then kept at 4–8°C for a fixed and previously established period of time (one day or 2, 12, 24, 36, 48 or 60 months) and submitted only once to repeated measurements of the same hormones at the end of storage timing.
Blood measurements were performed by immunofluorometric assays (AutoDELFIA™ Neonatal hTSH) or fluoroimmunossays (AutoDELFIA™ Neonatal Thyroxine and AutoDELFIA™ Neonatal 17α-OH-progesterone) in AutoDELFIA 1235 Automated Immunoassay Systems (Wallac Oy, Turku, Finland). Results were expressed in per unit of blood volume. Minimal detectable concentrations were 1.0 µU/mL for TSH, 1.0 µg/dL for T4 and 2.5 ng/mL for 17-OHP. Inter-assay variation was ≤4.9, 2.8 and 2.6% for TSH levels of 0.5, 4.6 and 23.1 µU/mL, respectively; ≤13.8, 11.0 and 16.4% for T4 levels of 1.3, 3.1 and 5.4 µg/dL, respectively; and ≤10.0 and 8.5% for 17-OHP levels of 8.3 and 20.4 ng/mL, respectively. Reference ranges for children born at term are ≤10.0 µU/mL for TSH, 3.5–11.0 µg/dL for T4 and ≤10.0 ng/mL for 17-OHP, as suggested by the manufacturer.
Statistical analysis
Basal hormonal measurements showed a normal distribution when all the 228 samples were tested as a single group using the Kolmogorov-Smirnov test (P > 0.05). Hormonal data are presented as mean and 95% CI. Differences between basal and repeated measurements in each group were evaluated by Student's paired t-test. Statistical analysis was performed using Minitab® software (version 15.1 for Windows, Minitab Inc., State College, PA, USA) and differences were considered significant at P < 0.05.
Mean (95% CI) blood thyroid-stimulating hormone (TSH) levels measured in filter paper 24–48 hours after collection (basal) and following one day or 2, 12, 24, 36, 48 or 60 months of sample storage at 4–8°C
RESULTS
Basal and repeated hormonal measurements in each group are compared in Tables 1–3. Blood T4 concentrations showed an unexpected rise at two months of sample storage, which was not confirmed at 12, 24 or 36 months (Table 2). Blood TSH and T4 levels declined significantly only when they were re-assayed at 48 and 60 months of sample storage (Tables 1 and 2). In contrast, 17-OHP concentrations decreased earlier, starting at 24 months and continuing throughout the remaining period (Table 3).
Mean (95% CI) blood thyroxine (T4) levels measured in filter paper 24–48 hours after collection (basal) and following 1 day or 2, 12, 24, 36, 48 or 60 months of sample storage at 4–8°C
Mean (95% CI) blood 17-hydroxyprogesterone levels (17-OHP) measured in filter paper 24–48 hours after collection (basal) and following 1 day or 2, 12, 24, 36, 48 or 60 months of sample storage at 4–8°C
DISCUSSION
Storage of neonatal screening samples is important to generate a biological specimen bank for hormonal testing as well as for studying alleles and other molecular markers of genetic diseases in population studies or retrospective projects. 8 Among the technical variables influencing the accuracy of screening tests, time and condition of sample storage represent an important but still unresolved issue. 9 Other factors include time of collection after birth, season, method of blood spotting as well as temperature, humidity, technical manipulation and time of elution from the filter paper. 7,9–11 In addition, lot-to-lot differences in the filter papers may also contribute to variability in neonatal testing, although previous studies have suggested that they alone are not sufficient to cause significant errors in neonatal screening. 11
In Brazil, clinical laboratories usually store neonatal filter paper samples at 4–8°C for one year. Our findings support this approach, since TSH and T4 concentrations dropped significantly only when samples were stored for periods longer than 36 months. Blood 17-OHP levels declined earlier, when samples were kept for more than 12 months.
Most specimens selected for this study had basal hormonal levels within the reference range due to the difficulty in finding samples with markedly abnormal results. However, we have no reason to believe that the latter would behave differently if collected, stored and analyzed under the same conditions described in this study.
Other limitations of this study include the small number of samples selected for each group and restricted sample availability, which did not allow repeated hormonal measurements in the same specimen throughout the entire 60-month period. These may explain the unexpected rise in T4 levels seen after two months of sample storage, which was not confirmed at subsequent times. In addition, specimen stability was only evaluated at 4–8°C and not at other conditions. It is possible that storage of samples at −20°C or in sealed containers with desiccants would improve the stability of specimens along time. Waite et al. 9 demonstrated that T3 and TSH are relatively stable for one month when samples are exposed to high humidity or high and variable temperatures (−20°C, 4°C, 25°C and 37°C). In contrast, T4 was most affected by these factors, decreasing significantly after four days of suboptimal storage conditions. 9 In regard to 17-OHP stability, Török et al. 12 demonstrated that this hormone declined 2–3% annually in dried blood spots kept at room temperature. Since their inter-assay variation was 8%, the authors suggested that this small decrease would not affect the interpretation of results when samples are measured after a few years of storage.
In summary, our data suggest that neonatal screening of filter paper samples kept at 4–8°C are reliable for repeating hormonal measurements when specimens are stored for up to one year, in the case of 17-OHP, or three years, in the case of T4 and TSH.
Footnotes
ACKNOWLEDGEMENTS
We thank the staff of Laboratório de Hormônios e Genética Molecular LIM/42 for their technical support on hormonal measurements.