ACCURACY OF DIFFERENT GH PROVOCATIVE TESTS FOR THE DIAGNOSIS OF GH DEFICIENCY IN CHILDREN

Program: Abstracts - Orals, Featured Poster Presentations, and Posters
Session: SUN 624-646-Growth: Clinical Trials & Observational Studies
Clinical
Sunday, June 16, 2013: 1:45 PM-3:45 PM
Expo Halls ABC (Moscone Center)

Poster Board SUN-638
Chiara Guzzetti*1, Anastasia Ibba1, Sabrina Pilia1, Nadia Beltrami2, Natascia Di Iorgi3, Alessandra Rollo4, Giorgio Radetti2, Stefano Zucchini4, Mohamad Maghnie5 and Sandro Loche1
1Ospedale Microcitemico, Cagliari, Italy, 2Ospedale Generale Regionale Bo, Bolzano, Italy, 3IRCCS G Gaslini, Genova, Italy, 4Univ Bologna, Bologna, Italy, 5UniversitÓ di Genova, IRCCS Giannina Gaslini, Genova, Italy
Objective:The diagnosis of GH deficiency (GHD) in children and adolescents is based on reduced peak GH response to at least two provocative tests. The cut-off limits of peak GH is arbitrarily set between 7-10 µg/l. The aim of this study was to evaluate the diagnostic accuracy of different cut-off levels for the GH response to Clonidine (Clo), Insulin Tolerance Test (ITT), Arginine (Arg) and Spontaneous Sleep (SS).

Patients and methods:This was a retrospective study in 629 short children and adolescents (402 M and 227 F, aged 2-17.8 yr) who underwent GH secretion studies for short stature. All underwent at least one test that was repeated if GH peak was <10 µg/L. IGF-I was also measured in all children. GHD was diagnosed in children with two GH peaks <10 µg/L (192 GHD and 400 controls). GH and IGF-I were measured by chemiluminescence assay in all samples. ROC analysis was used to evaluate the diagnostic accuracy of the tests. ROC analysis was repeated also assuming 7 µg/L as a cut-off for the diagnosis of GHD (111 GHD and 518 controls).

Results: ROC analysis showed that at a cut-off levels of 7 µg/L the diagnostic accuracy was better for Arg (AUC=0.9, Sensitivity=100%, Specificity=72.6%, vs AUC=0.87, Sensitivity=100%, Specificity=63.1%), ITT (AUC=0.9, Sensitivity=100%, Specificity=62.6%, vs AUC=0.87, Sensitivity=100%, Specificity=49.7%) and SS (AUC=0.96, Sensibility=100%, Specificity=88.9%, vs AUC=0.93, Sensitivity=100%, Specificity=84.3%). The diagnostic accuracy of Clo was better at a cut-off of 10 µg/L (AUC=0.99, Sensibility=100%, Specificity=97.2%, vs AUC=0.98, Sensitivity=100%, Specificity=89.8%). ROC analysis showed that IGF-I has low accuracy in the diagnosis of GHD using both cut-offs (AUC=0.66 at a cut-off of 7µg/L and AUC=0.61 using peak GH cut-off of 10 µg/l).

Conclusions: The use of peak GH cut-off limit of 7 µg/l improves the diagnostic accuracy of Arg, ITT and SS while the diagnostic accuracy of Clo was better at a cut off of 10 µg/L. SS showed the best accuracy at both cut-offs. IGF-I is characterized by low diagnostic accuracy. These results suggest that the choice of a cut-off limit of the GH peak for the diagnosis of GHD might be dependent on the type of stimulation test.

Nothing to Disclose: CG, AI, SP, NB, ND, AR, GR, SZ, MM, SL

*Please take note of The Endocrine Society's News Embargo Policy at http://www.endo-society.org/endo2013/media.cfm