Short stature and SHOX (Short stature homeobox) variants-efficacy of screening using various strategies

Abstract

Background: SHOX mutations have previously been described as causes of Léri-Weill dyschondrosteosis (LWD), Langer mesomelic dysplasia (LMD), and idiopathic short stature. The loss of X chromosome-Turner syndrome or mosaic 45,X/46,XX or 46,XY-also leads to the heterozygous loss of SHOX in patients with short stature only or with features similar to LWD. The aim of this study was to assess the efficacy of the targeted screening for SHOX variants, which involved different methods in the laboratory analysis of short stature. We determined the significance and positive predictive value of short stature for the detection of SHOX variants.

Methods: Targeted screening for variants in SHOX involving MLPA, sequencing, karyotyping and FISH was performed in the short stature cohort (N = 174) and control cohort (N = 91). The significance of short stature and particular characteristics for the detection of SHOX variants was determined by Fisher's exact test, and the probability of SHOX mutation occurrence was calculated using a forward/stepwise logistic regression model.

Results: In total, 27 and 15 variants influencing SHOX were detected in the short stature and control cohorts, respectively (p > 0.01). Sex chromosome aberrations and pathogenic CNV resulting in diagnosis were detected in eight (4.6%) and five (2.9%) patients of the short stature group and three (3.3%) and one (1.1%) individuals of the control group. VUS variants were discovered in 14 (8.0%) and 11 (12.1%) individuals of the short stature and control groups, respectively. MLPA demonstrated the detection rate of 13.22%, and it can be used as a frontline method for detection of aberrations involving SHOX. However, only mosaicism of monosomy X with a higher frequency of monosomic cells could be reliably discovered by this method. Karyotyping and FISH can compensate for this limitation; their detection rates in short stature group were 3.55% and 13.46% (N = 52), respectively. FISH proved to be more effective than karyotyping in the study as it could reveal cryptic mosaics in some cases where karyotyping initially failed to detect such a clone. We suggest adding FISH on different tissue than peripheral blood to verify sex-chromosome constitution, especially in cases with karyotypes: 45,X; mosaic 45,X/46,XX or 46,XY; 46,Xidic(Y) detected from blood; in children, where mosaic 45,X was detected prenatally but was not confirmed from peripheral blood. The correlation of short stature with the occurrence of SHOX mutations was insignificant and short stature demonstrates a low positive predictive value-15.5% as unique indicator for SHOX mutations. The typical skeletal signs of LWD, including Madelung deformity and disproportionate growth, positively correlate with the findings of pathogenic SHOX variants (p < 0.01) by Fisher's exact test but not with the findings of VUS variants in SHOX which are more prevalent in the individuals with idiopathic short stature or in the individuals with normal height.

Keywords: FISH; Idiopathic short stature; Karyotyping; Leri-Weill dyschondrosteosis; MLPA; SHOX; Screening for mutations; Sequencing; Short stature; Turner syndrome.

Figure 1. Flowchart of the performed investigations in short stature patients.

(blue arrows—routine testing, red arrows—additional testing in case of a pathogenic finding).

Figure 2. Frequency of the particular SHOX aberrations in individuals of the short stature (A) and control cohort (B).

Figure 3. Contribution of the particular methods to the detection of SHOX variants in the short stature cohort.

The number of detected CNV for the particular methods in the short stature cohort N = 27. The detection of pathogenic findings was made by three methods in 5 samples* and by two methods in one sample**. However, each method contributes differently to clarify patients’ phenotype. We were able to detect different clone by FISH on a buccal smear (47,XXX in 1 out of 5 samples or 45,X in 1 out of 5 samples). In one sample** we detected duplication of SHOX by MLPA but 45,X clone was discovered by FISH on the buccal smear. By FISH 1 case (mosaic of 45,X) was exclusively detected. The method helped to specify altogether the findings in 3 cases (arrows) and the result concordant with MLPA and karyotyping was achieved in further 3 cases by FISH. (The detection rate for the particular methods related to the number of performed tests is stated in the text).

Figure 4. CNV detected by MLPA and Sanger sequencing in the short stature and control cohort.

CNV detected by MLPA and Sanger sequencing in the short stature and control cohort. Diagnosis in the patients is marked with an asterisk, ISS patients are without an asterisk.

Figure 5. Family tree in (A) duplication of the upstream enhancers elements (B) deletion of the downstream enhancers elements.

The family tree in (A) duplication of the single upstream enhancer element CNE3 in propositus with ISS and cleft palate (Table 1, Table S1) (B) recurrent deletion of the downstream enhancers elements - 47.5 kb CNE7-CNE9 in individuals of the control group (Table 1, Table S1) (blue symbols—detected variant).