Endocrine Abstracts

JOINT1847

Introduction: Short-stature homeobox-containing (SHOX) gene haploinsufficiency is the most frequent monogenic cause of short stature and is responsible for highly variable phenotypes ranging from idiopathic short stature (ISS) to severe dyschondrosteosis in the Langer syndrome. The reported prevalence of SHOX haploinsufficiency in children with apparently ISS varies from 2 to 17%, and its radiological expression is variable. The aim of this study was to find out whether simple radiological characteristics on the left-hand and wrist radiography were associated with the presence of SHOX gene variations and could be a good indicator for SHOX gene molecular study.

Materials and Methods: This study is a descriptive, retrospective and non-interventional study based on clinical and radiological data of 266 patients with apparently ISS referred to the Pediatric Endocrinology Unit of Angers University hospital from 2016 to 2023 for whom SHOX analysis was performed by Multiplex Ligation-dependent Probe Amplification (MLPA) and/or sequencing. To further analyze the sensitivity of radiological criteria for the molecular diagnosis of SHOX variations, we additionally included 33 patients diagnosed with a SHOX mutation between 2010 and 2015.

Results: 11% of patients with apparently ISS had a SHOX deficiency. SHOX deficiency was more often associated with convexity of the distal radial metaphysis (P < 0, 001), pyramidalisation of the carpal row (P < 0, 001) and lucency of the ulnar border of the distal radius (P < 0, 001), but no difference was found considering the triangularization of the distal radial epiphysis (P = 0, 142). A cut-off point of 148 degrees for the convexity of the distal radial metaphysis had a sensitivity of 90% and a specificity of 47, 8%, and a cut-off point of 129 degrees for the pyramidalisation of the carpal row had a sensitivity of 87, 5% and specificity of 50, 3%.

Conclusion: We found cutoffs for SHOX gene variation diagnosis from the simple left-hand and wrist radiology. These cut-offs had a fair sensitivity (90%) and a mild specificity (50%). This suggests that in structures where the resources for SHOX molecular study are limited, an oriented screening based on these cutoffs would decrease the number of molecular studies by half (and would miss 10% of the SHOX gene variations), whereas, in structures where molecular studies are easily available, systematic screening of children with otherwise unexplained short stature might be useful.