ORF03-05 - Predictors of a Genetic Diagnosis in Short Stature: A Real-World Pediatric Cohort

Background: Genetic evaluation is increasingly integrated into the assessment of pediatric short stature (SS), but real-world data on diagnostic yield, phenotypic predictors, and biochemical patterns remain limited.

Objective: To assess genetic causes of SS and identify features associated with a genetic diagnosis in a real-world pediatric cohort.

Methods: Children referred for SS were retrospectively evaluated. Inclusion criteria were height ≤−2 SDS or below parental target range. Patients with typical growth hormone (GH) deficiency, Turner syndrome, or constitutional delay of growth and puberty were excluded. Genetic testing was phenotype-driven and included chromosomal microarray, imprinting disorder testing, targeted SS/skeletal dysplasia gene panels, and whole-genome sequencing. Patients were classified into diagnostic groups based on gene function, including GH axis-related, bone-related, non-GH/non-bone related, variants of uncertain significance (VUS), pre-referral genetic diagnoses, atypical GH deficiency; those without an identified genetic diagnosis were divided into small for gestational age (SGA) and non-SGA subgroups. IGF-1 concentrations were categorized relative to age- and sex-specific reference ranges. Statistical analyses included ANOVA and multivariate logistic regression.

Results: We included 193 patients (113 M). At referral, boys were aged 6.37±4.65 years and girls 6.61±4.17 years; mean height SDS was -2.96±1.27 and -3.10±1.57, respectively. A genetic diagnosis was identified in 36.8%, with an additional 5.7% carrying VUS. Bone-related genetic variants were most common, including in patients with pre-referral diagnoses (52.4%), and were associated with more severe SS and higher BMI SDS than other groups (p < 0.01). Genetic diagnoses were found in children with height ≤−2.5SDS (42.2%), in those with height >−2.5 to ≤−2SDS (32.6%) and in children with height >−2SDS, short for parental height (60%), 28% of whom had a pre-referral genetic diagnosis. Head circumference (HC) patterns differed across groups; both microcephaly and normocephaly associated with increased likelihood of genetic diagnoses. Multivariate analysis identified dysmorphic features and BMI SDS as strong independent predictors of genetic diagnoses, with height SDS and HC SDS contributing additional predictive value. High IGF-1 concentrations were observed in 25% of patients with bone-related variants, low IGF-1 concentrations were more frequent in atypical GH deficiency.

Conclusions: Genetic testing has high yield across pediatric SS. Bone-related disorders are most common and show typical auxological and biochemical profiles, including high IGF-1. Integrating clinical predictors such as dysmorphic features, BMI SDS, and HC into diagnostic algorithms may improve genetic test selection and enable more personalized management.