ESPE Abstracts (2019) 92 P1-72

Diagnostic Value of Random Serum Growth Hormone (GH), IGF-I and IGFBP-3 Concentrations for the Diagnosis of Growth Hormone Deficiency (GHD) in Patients Below One Year of Life

María Gabriela Ballerini, Débora Braslavsky, Analía Verónica Freire, Ana Keselman, María Eugenia Rodríguez, Mercedes Altube, Paula Alejandra Scaglia, Ignacio Bergadá, María Gabriela Ropelato


Centro de Investigaciones Endocrinológicas ‘Dr. César Bergadá’ (CEDIE)- CONICET – FEI – División de Endocrinología-Hospital de Niños Ricardo Gutiérrez, Buenos Aires, Argentina


GHD diagnosis in neonates and infants is a challenge owing to the fact that GH pharmacological stimulation tests (GHST) are not approved at this age. In a retrospective study, we found that a random GH<6.5 µg/L confirmed GHD diagnosis in neonates with clinical suspicion of GHD with high diagnostic accuracy (1). The accuracy of GH and its surrogates of action have not been set for infants with current standardized immunoassays.

Objective: 1-To evaluate serum GH/IGF-I/IGFBP-3 concentrations in healthy neonates/infants in order to obtain normal reference intervals. 2-To investigate GH/IGF-I/IGFBP-3 usefulness for GHD diagnosis in infancy.

Design: Diagnostic validation study. Tertiary public hospital.

Methods: 430 healthy neonates/infants (2days-1year, 184 girls); 120 patients referred with clinical suspicion of GHD (2days-1year, 43 girls) from March 2015 to March 2019. Gold standard for GHD diagnosis: growth retardation, other pituitary hormone deficiencies, brain MRI abnormalities or GHST during childhood. Infants in whom GHD was ruled out were diagnosed as having congenital hyperinsulinism (HI) or non-GHD. Reference values obtained in healthy neonate/infants were used to calculate SDS. Main outcomes by ROC were: sensitivity(S), specificity(Sp), negative predictive values(NPV) and area under the curve(AUC) of GH/IGF-I/IGFBP-3 (IMMULITE 2000XPi-Siemens).

Results: Controls: the highest GH and the lowest IGF-I and IGFBP-3 concentrations were found in 1st week old neonates. Thereafter, GH decreased and IGF-I/IGFBP-3 increased with age to reach the lowest GH and the highest IGF-I/IGFBP-3 concentrations beyond 6 months of age. GH and IGFBP-3 were always detectable in infants while 13% presented non-detectable IGF-I.

Patients: GHD was diagnosed in 41, HI in 24 and 56 were non-GHD. Only GH-SDS and IGF-I-SDS were significantly lower in GHD than non-GHD or HI (P<0.0001). In infants, a GH cut-off of <–1.3 SDS presented: S: 0.94(0.74–0.99), Sp: 0.67(0.56–0.77), NPV: 0.98(0.90–1.0) and AUC: 0.90(0.83–0.95). IGF-I-SDS cutoff of <–1,1 SDS had: S: 0.73(0.58–0.85), Sp: 0.74(0.61–0.84), NPV: 0.79(0.68–0.89) and AUC: 0.75(0.67–0.84).

Conclusions: GH/IGF-I/IGFBP-3 concentrations should be weighed in the context of the clinical presentation and MRI findings. However, this study shows that GH constitutes a highly reliable biomarker for GHD screening in infants with clinical suspicion of GHD due to its high sensitivity and NPV. These data, in addition to our previous results on neonates, suggest that a random GH could be useful for GHD diagnosis throughout the entire first year of life.

(1)Ballerini et. al. Horm Res Paediatr 2018