RÉSUMÉ
Objective:To establish the cut-off value of tetradecenoyl carnitine (C14∶1)/dodecenoyl carnitine(C12∶1) based on non-derivatized tandem mass spectrometry (MS/MS), and to explore the application value of C14∶1/C12∶1 to screen newborns for very long chain acyl-CoA dehydrogenase deficiency (VLCADD), determining the best combination of indicators for screening VLCADD.Methods:This retrospective study included data from 17 newborns with VLCADD detected by MS/MS and confirmed by acyl-CoA dehydrogenase very long chain ( ACADVL) gene detection, and 423 507 newborns with normal MS/MS results. The data from these newborns were collected from January 2014 to December 2021 as the newborns received neonatal screening in Nanjing Neonatal Disease Screening Center and Suzhou Neonatal Disease Screening Center. All newborns were divided into 3 groups: all newborns group, full-term newborns group and normal-birth-weight newborns group, and the cut-off values of C14∶1/C12∶1 for VLCADD in these 3 groups were determined by their receiver operating characteristic (ROC) curves individually. With these results, a total of 5 interpretation schemes were composed using different indicators alone or jointly: scheme 1 being C14∶1/C12∶1, scheme 2 being C14∶1, scheme 3 being C14∶1+C14∶1/C2+C14∶1/C16, scheme 4 being C14∶1/C12∶1+C14∶1, and scheme 5 being C14∶1/C12∶1+C14∶1+C14∶1/C2+C14∶1/C16. The detection rate, false-positive rate and positive predictive value of each scheme were calculated, and their screening efficiencies were statistically compared by Chi-square tests. Results:The cut-off values of C14∶1/C12∶1 for VLCADD in the 3 newborn groups were all 2.80. The detection rates of VLCADD with all 5 interpretation schemes were 17/17. Scheme 1 had the highest false positive rate [26.15‰ (11 075/423 524)] and the lowest positive predictive value [0.15% (17/11 092)]. Scheme 4 (Scheme 5) had the lowest false positive rate [0.02‰ (10/423 524)] and the highest positive predictive value [62.96% (17/27)]. Comparing scheme 4 (Scheme 5) with scheme 1, scheme 2 and scheme 3, the differences of false positive rate (χ2=302.30,11 191.50,32.06) and positive predictive value (χ2=102.51,3 485.61,13.83) were statistically significant (all P<0.001). Conclusion:C14∶1/C12∶1 was an effective auxiliary interpretive indicator for VLCADD in newborn screening, and the combination of C14∶1/C12∶1+C14∶1 was tested to be the best indicator for VLCADD screening based on non-derivatized tandem mass spectrometry.
RÉSUMÉ
<p><b>OBJECTIVE</b>To investigate the correlation of gestational age and birth weight with 17α-hydroxyprogesterone (17α-OHP) levels, and with results of adrenal hyperplasia newborn screening.</p><p><b>METHOD</b>Using time-resolved fluorescence immunoassay, the authors measured concentrations of heel blood 17α-OHP by newborn dried blood spots on filter paper which included 29 hospitals newborns of Wujiang, Taicang, Zhangjiagang, Kunshan, and Suzhou, where there were 118 050 infants in total who had accurate gestational age and birth weight (62 490 males, 55 560 females). According to the classification by gestational age, there were 4 693 premature infants, 113 300 term infants and 57 overdue infants. According to the classification by birth weight, there were 4 172 infants with weight < 2 500 g, and 113 878 infants weight ≥ 2 500 g. And, in all premature infants, gestational age of 189 infants was < 32 weeks, 2 277 infants less than 36 weeks but ≥ 32 weeks, and 2 227 infants less than 37 weeks but not less than 36 weeks. Neonatal heel blood concentration of 17α-OHP was measured by dissociation enhanced lanthanide fluorescence immunoassay (DELFIA), and the correlation between 17α-OHP and gestational age or birth weight was retrospectively analyzed by using Spearman test.</p><p><b>RESULT</b>The distribution of 17α-OHP levels was skew. The 17α-OHP levels decreased significantly from very preterm births, moderately preterm, later period preterm to term infants [19.21 (8.07, 24.00), 12.35 (6.81, 18.00), 8.58 (5.66, 13.80), 5.60 (3.57, 8.51) , 3.34 (2.58, 5.23) nmol/L; 479.42, 62.25, 36.24, 23.30, 13.73 nmol/L;P all = 0.000]. The 17α-OHP levels decreases from very low birth weight (VLBW), extremely low birth weight (ELBW), low birth weight (LBW), normal birth weight to macrosomia [5.24 (3.24, 8.96) , 11.30 (6.84, 22.95) , 8.50 (5.28, 14.90) , 5.66 (3.61, 8.62) , 5.38 (3.40, 8.11) nmol/L; 485.26, 125.18, 39.50, 23.80, 22.15 nmol/L; P = 0.000 for all comparison]. Neonatal 17α-OHP levels and gestational age, body weight was significantly negatively correlated respectively -16.40 and -10.10 (P both = 0.000) by using Spearman test. Neonatal 17α-OHP levels and gestational age, body weight were binomially distributed, and the formulae were y = 0.105 5x²-2.457 6x + 17.689, R² = 0.980 3 and y = 0.411x²-3.988x+14.75, R² = 0.983. Little preterm infants, preterm infants and term infants in low birth weight infants 17α-OHP levels were significantly higher than non-low birth weight infants [11.20 (6.01, 18.90) vs 9.05 (5.85, 14.90) nmol/L, 9.76 (4.32, 10.35) vs 5.59 (3.56, 8.48) nmol/L, P all = 0.000].</p><p><b>CONCLUSION</b>Neonatal 17α-OHP levels and gestational age, body weight was significantly negatively correlated; in order to improve the accuracy and sensitivity, cut-off value of neonatal 17α-OHP should be adjusted according to gestational age and weight.</p>