This study was approved by the Institutional Review Board (2023-01-017-000) of Inha University Hospital. We reviewed the medical records of 50 patients who visited our pediatric endocrinology clinics from June 2019 to December 2022 and underwent genetic testing for suspected genetic short stature.

The exclusion criteria were as follows: short stature caused by (1) chromosomal aneuploidy or abnormalities (such as Down syndrome, Turner syndrome, chromosome 22q11.2 deletion); (2) multiple pituitary hormone deficiency (MPHD); (3) chronic systemic disorder such as chronic renal failure, chronic inflammatory disease, or inherited metabolic disorder; and (4) growth failure secondary to psychological factors.

We enrolled 37 patients from 35 families in this study. We obtained written informed consent from all parents or gardians before genetic testing using NGS. In all cases, conventional methods such as karyotyping and single gene sequencing failed to determine the genetic cause of short stature. All enrolled patients met one or more of the following criteria: (1) family history of short stature with similar phenotypes; (2) severe short stature, defined as a height-standard deviation score (Ht-SDS) of less than -3; (3) additional congenital anomalies or facial dysmorphisms; (4) evidence of skeletal deformities; and (5) associated intellectual disability (ID) or developmental delay (DD).

We assessed family history through three generations, including parental height for the calculation of midparental height (MPH), and evaluated the patient's detailed clinical phenotypes. If skeletal deformity was suspected, a radiological skeletal survey was performed.

Height and weight are expressed as standard deviation scores (Ht-SDS and Wt-SDS) and were calculated based on growth standards for Korean children and adolescents [14]. MPH was calculated as follows: for females, (maternal height + paternal height − 13 cm)/2, and for males, (maternal height + paternal height + 13 cm)/2. Small for gestational age (SGA) was defined as birth weight below the third percentile of a standard sexspecific birth weight curve for gestational age in South Korea [15]. Insulin-like growth factor-1 (IGF-1) SDS at the first visit/referral or before growth hormone (GH) treatment was calculated according to the Korean standard reference [16]. GH deficiency (GHD) was confirmed, via 2 different GH stimulation tests using insulin, arginine, or L-dopa, as a peak GH response of <10 ng/mL in both tests.

In this study, we administered 1 of 3 genetic tests (targeted panel sequencing using 1 of 2 different panels or WES) to each patient according to their phenotypes and accompanying symptoms. After obtaining written informed consent from parents or guardians, genomic DNA was extracted from the peripheral blood of the patients and, where possible, their parents and/or siblings. Among the 37 enrolled patients, 12 underwent panel sequencing using a short stature-targeted panel (129 genes, Supplementary Table 1), 7 patients with suspected skeletal deformity underwent panel sequencing using a skeletal dysplasia-targeted panel (433 genes, Supplementary Table 2), and 18 patients underwent WES.

For the targeted panel sequencing, all exons of the target genes were captured using a Celemics custom panel (Celemics, Seoul, Korea) or an Integrated DNA Technologies (IDT) custom panel (IDT, Coralville, IA, USA). Sequencing was performed on the Illumina NextSeq platform (Illumina, San Diego, CA, USA) and generated 2×150 bp paired-end reads. DNA sequence reads were aligned to the reference sequence based on the public human genome build UCSC GRCh37/hg19. Alignment was performed with BWA-MEM, and duplicate reads were marked with biobambam2. Base quality recalibration and variant calling were performed with the Genome Analysis Toolkit, and annotation was performed with Variant Effect Predictor and dbNSFP. Sequence variants were classified using the ACMG/ AMP guidelines [17].

For WES, genomic DNA was extracted from proband blood. All exon regions of all human genes (~22,000) were captured by a Twist Human Core Exome kit (Twist Bioscience, South San Francisco, CA, United States). The captured regions of the genome were sequenced using a Novaseq 6,000 sequencing machine (Illumina). The raw genome sequencing data analysis, including alignment to the GRCh37/hg19 human reference genome, variant calling, and annotation, was conducted using open-source bioinformatics tools and in house software. We extracted evidence data on the pathogenicity of variants from previous studies and disease databases, particularly ClinVar (https://www.ncbi.nlm.nih.gov/clinvar/) and UniProt (https://www.uniprot.org/). For the screening and reading of genetic mutations, EVIDENCE v.3.218)), an automatic mutation analysis system developed by 3billion, was used. Sequence variants were classified using the ACMG/AMP guidelines [17,19,20].

All suspected pathogenic/likely pathogenic (P/LP) variants were confirmed by Sanger sequencing, and copy number variants were identified by chromosomal microarray (CMA). According to the inheritance patterns of the disease, results were considered positive if one or 2 P/LP variants were identified in the disease-related genes. We report cases without a variant of uncertain significance or a P/LP variant as negative results.

Fisher exact test and the Wilcoxon rank-sum test were performed between the groups. Results with P<0.05 were considered statistically significant. All analyses were performed using IBM SPSS Statistics ver. 26.0 (IBM Co., Armonk, NY, USA).

This study was performed according to the Helsinki Declaration (http://www.wma.net/en/30publications/10policies/b3/) and was approved by the Institutional Review Board of Inha University Hospital (2023-01-017-000). Written informed consent was obtained from all subjects.

We included 37 Korean patients (17 females and 20 males) from 35 nonconsanguineous families in this study. Their median age at diagnosis was 10.5 years (range, 6 months to 17.5 years), and their mean Ht-SDS at diagnosis was -3.00±0.85. Patient demographics, auxological profiles, and clinical characteristics are listed in Table 1. Severe short stature (18 of 37, 48.6%) and facial dysmorphism (17 of 37, 45.9%) were the most common clinical manifestations. Other clinical manifestations included skeletal deformities (13 of 37, 35.1%), ID/DD (13 of 37, 35.1%), SGA (11 of 37, 29.7%), family history (10 of 37, 27%), and congenital anomalies (6 of 37, 16.2%). The 37 patients with various clinical features were divided into 2 groups, one containing those with congenital anomalies, facial dysmorphism, skeletal deformities, or ID/DD (syndromic group), and one containing those with only a family history or severe short stature (nonsyndromic group). Their diagnostic rates and clinical characteristics are compared in Table 1. The basal characteristics did not differ significantly between the groups. The diagnostic rate was higher in the syndromic group than the nonsyndromic group (46.1% vs. 27.2%, P<0.01) (Table 1).

Genetic analyses using NGS and further validation studies (Sanger sequencing or CMA) identified 13 causative genes in 15 patients (7 females, 8 males). The overall diagnostic yield was 40.5%, with the highest diagnostic yield of 57.1% for targeted skeletal dysplasia panel sequencing, 41.6% for targeted short stature panel sequencing, and 27.7% for WES. The genetic variants of the 15 patients with confirmed clinical and genetic diagnoses are summarized in Table 2. The identified genes and final diagnoses of those patients were as follows: 2 Noonan syndrome, (MIM#164790 and MIM#163950) caused by an NRAS variant and PTPN11 variant, respectively; 2 hypochondroplasia (MIM#146000) by FGFR3 variants; 2 Seckel syndrome 5 (MIM#3613823) in siblings by CEP152 variants; short stature and advanced bone age, with or without early-onset osteoarthritis and/ or osteochondritis dissecans (MIM#165800) by an ACAN variant; KBG syndrome (MIM#148050) by an ANKRD11 variant; Coffin-Siris syndrome 1 (MIM#135900) by an ARID1B variant; osteogenesis imperfecta type 3 (MIM#259420) by a COL1A2 variant; metaphyseal chondrodysplasia, Schmid type (MIM#120110) by a COL10A1 variant; exostoses, multiple, type 1 (MIM#133700) by an EXT1 variant; Cornelia De Lange syndrome 1 (MIM#122470) by an NIPBL variant; Leri-Weill dyschondrosteosis (MIM# 127300) by a SHOX variant; and X-linked Allan-Herndon-Dudley syndrome (MIM#300523) by an SLC16A2 variant. One gene (CEP152) was inherited in an autosomally recessive manner, 1 gene (SLC16A2) was inherited in an X-linked manner, 1 gene (SHOX) was inherited in a pseudo-autosomally dominant manner, and the remaining 10 genes were inherited in an autosomally dominant manner.

Table 3 shows the clinical characteristics and phenotypes of the positive and negative groups. The basal characteristics and auxological profiles did not differ significantly between them. In the positive group, only 2 patients (13.3%) received GH stimulation tests, and neither of them was diagnosed with GHD. On the other hand, most of the negative group received GH stimulation tests, and 64.7% of them were diagnosed with GHD. Among the major clinical features, SGA (46.6% vs. 18.1%, P=0.031) and skeletal deformity (53.3% vs. 22.7%, P=0.027) occurred more frequently in the positive group. Other clinical characteristics did not differ significantly between the groups (Table 3).