ESPE Abstracts

In a retrospective study (1), we found that a GH<6.5 μg/L, IGF-I-WHO87/518 <30 μg/L and IGFBP-3<0.8 μg/mL confirmed GHD diagnosis with high diagnostic accuracy in neonates with clinical suspicion of GHD. GH and insulin negatively regulate IGFBP-2, and it was proposed to reflect GH status in the diagnostic work-out of GHD in childhood and adults. The accuracy of IGFBP-2 has not been set for neonates.

Objective: To prospectively validate own GH/IGF-I/IGFBP-3 cut-off values and to investigate the usefulness of IGFBP-2 for diagnosing GHD in neonates.

Methods: The study included 16 neonates that were referred with clinical suspicion of GHD from March 2017 to March 2018 and 18 control neonates (median: 21 days). GHD diagnosis was based on growth retardation, other pituitary hormone deficiencies or brain MRI. Neonates in whom GHD was ruled out were diagnosed as congenital hyperinsulinism or with transient disorders (non-GHD). A new cut-off for current IGF-I-WHO 02/254 by Siemens was calculated (26 ng/mL). Main outcomes (ROC): sensitivity, specificity, negative and positive predictive values (NPV; PPV) of IGFBP-2 (ELISA-MyBioSource).

Results: GHD was diagnosed in 6 neonates; hyperinsulinism in 3 and 7 had transient disorders. GH was <6.5 μg/L in 6/6 GHD and 2/8 non-GHD patients. IGF-I was <26 μg/L in 4/6 GHD, 1/7 non-GHD and 2/3 neonates with hyperinsulinism (undetectable IGF-I concentrations in 5/17 neonates). GH was significantly lower in GHD than the other groups (P<0.001). No differences were obtained for IGF-I or IGFBP-3. GH and IGFBP-2 presented an inverse correlation (r=−0.79; P<0.01). IGFBP-2 concentration was significantly higher in GHD (median: 314 μg/L) compared to non-GHD (97 μg/L) or hyperinsulinism groups (68 μg/L) (P<0.001). Most non-GHD and hyperinsulinemic neonates had IGFBP-2 concentrations within control range (210 μg/L). IGFBP-2 outcomes by ROC (cut-off: 230 μg/L) were sensitivity: 1.0; specificity: 0.90; PPV: 0.83; NPV: 1.0 (P=0.002). 1/2 non-GHD that failed to reach GH cut-off presented IGFBP-2 <230 μg/L.

Conclusions: This study reaffirms that GH >6.5 μg/L excludes GHD with high diagnostic accuracy. IGF-I seems to be less useful, probably due to methodological sensitivity. The inclusion of IGFBP-2 strengthens GHD diagnosis. A larger sample size should be study to further consider IGFBP-2 measurement as a reliable biomarker to rule out GHD in neonates. Reference: (1) Horm Res in Paediatrics 2017;P2-804.