Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2024 Jun 6;56(11):1706-1710.
doi: 10.3724/abbs.2024089.

Significant biomarkers for predicting 1-month changes in IGF-1 in growth hormone-deficient children following r-hGH therapy

Fei Liu  1   2 , NokI Lei  3 , Wunying Li  3 , Yiwen Zheng  4 , Liangjian Hu  4 , Ronggui Hu  5 , Wenli Lu  3 , Yu S Huang  1   2   6
Affiliations

Significant biomarkers for predicting 1-month changes in IGF-1 in growth hormone-deficient children following r-hGH therapy

Fei Liu et al. Acta Biochim Biophys Sin (Shanghai). .
No abstract available

PubMed Disclaimer

Conflict of interest statement

Merck KGaA performed a fair-balanced review of the manuscript. The authors declare that they have no conflict of interest.

Figures

None
Figure 1
Identification of predictive factors for ΔIGF-1 through correlation, association and modelling analyses 1. Correlations between clinical variables and ΔIGF-1. (A) Clustering of all clinical variables by t-SNE with perplexity=3. ΔIGF-1 is colored in red. Four clinical variables, BMI, baseline IGF-1, baseline GH peak, and baseline IGFBP3, which were significantly associated with ΔIGF-1 according to univariate regression, are in bold. (B) Correlation heatmap between baseline clinical variables and ΔIGF-1. BMI, baseline IGF-1, baseline IGFBP3, and baseline GH peak were the four variables most strongly correlated with ΔIGF-1. BMI was positively correlated (green), while baseline IGF-1, IGFBP3, and GH peak were negatively correlated (brown). Two pairs, baseline IGF-1 (baseline IGFBP3) and baseline IR (baseline insulin), showed a strong correlation, which explains why baseline IGFBP3 was rendered entirely redundant by baseline IGF-1. The baseline GH peak was moderately correlated (r2≈0.55) with both the baseline IGF-1 and baseline IGFBP3 levels, which was not significant according to multivariate regression. 2. Significant SNPs correlated with ΔIGF-1 corrected by covariates. (C) Manhattan plot of 40 SNPs (*P<0.05). (D) LocusZoom plots of 4 SNPs (**P<0.01). The purple rhombuses represent significant SNPs, and the purple circles are two covariates. 3. Variable selection for the final prediction model by bootstrapping. (E) Each dot represents a variable. The x-axis is the mean effect size of a variable, and the y-axis is the frequency of nonzero effect sizes in 100 bootstrap runs. The larger the mean effect size is, the more likely the variable will be selected. (F) The x-axis is the number of variables. The y-axis is the frequency of nonzero effect sizes in 100 bootstrap runs. Most variables (the longest bar with frequency=1.0) are consistently selected, which shows model consistency. (G) The distribution of ρ (Spearman’s rank correlation) between the true and predicted values for 100 bootstrap runs; the median value of this ρ is 0.957. (H) The prediction results obtained by the final prediction model for one set of validation subjects (not included in the model training) achieved a ρ of 0.935.
None
Figure 2
Identification of potential biomarkers and pathways associated with GHD treatment (A) Clustering heatmap showing that the top 50 predictive probes (Supplementary Table S6) and 2 clinical variables can be used to distinguish between high-response and low-response subjects. High response: ΔIGF-1≥Q3; low response: ΔIGF-1≤Q1. (B‒E) Partial least squares-discriminant analysis (PLS-DA) plot demonstrating the applicability of using all probes in prediction models, showing very clear clustering of GHD patients’ r-hGH therapy response regarding the different levels of ΔIGF-1. (F) Volcano plot for differential gene expression probes in different response groups. Scattered points represent expression probes: the x-axis is the log2-fold change in the ratio of r-hGH-treated probes of patients vs baseline patients, whereas the y-axis is the ‒log10 P value, indicating the probability that a probe has statistical significance in its differential expression. The red dots represent probes that were significantly overexpressed after treatment, and the blue dots represent probes that were significantly downregulated after treatment. The probes with significant fold changes in the high-response group showed a strong association with the GH-IGF-1 axis, indicating that these probes potentially promote the GHD treatment response. (G‒K) Violin plots showing the associations between SNP genotypes and expression probes (FDR<0.05). The x-axis is the SNP genotype. The y-axis is the distribution of the expression values of the expression probes. The most significant eQTL association was between SNP rs12882504 in the intronic region of SOS2 and the ex-pression probe 237640_at of VCPKMT, which are relatively close in position on the chromosome, indicating that VCPKMT is a possible gene that affects the GHD response and that SOS2 may have a regulatory function. ns: not significant; **P<0.01.
See this image and copyright information in PMC

References

    1. Coste J, Letrait M, Carel JC, Tresca JP, Chatelain P, Rochiccioli P, Chaussain JL, et al. Long term results of growth hormone treatment in France in children of short stature: population, register based study. BMJ. . 1997;315:708–713. doi: 10.1136/bmj.315.7110.708. - DOI - PMC - PubMed
    1. Lu W, Shen S, Luo X, Gong C, Gu X, Li Y, Du M, et al. Comparative evaluation of short-term biomarker response to treatment for growth hormone deficiency in Chinese children with growth hormone deficiency born small for or appropriate for gestational age: a randomized phase IV open-label study. Ther Adv Endocrinol Metab. . 2013;4:41–49. doi: 10.1177/2042018813484051. - DOI - PMC - PubMed
    1. Mullis PE, Deladoëy J, Dannies PS. Molecular and cellular basis of isolated dominant-negative growth hormone deficiency, IGHD type II: insights on the secretory pathway of peptide hormones. Horm Res. . 2002;58:53–66. doi: 10.1159/000064663. - DOI - PubMed
    1. Stevens A, Murray P, Wojcik J, Raelson J, Koledova E, Chatelain P, Clayton P, et al. Validating genetic markers of response to recombinant human growth hormone in children with growth hormone deficiency and Turner syndrome: the PREDICT validation study. Eur J Endocrinol. . 2016;175:633–643. doi: 10.1530/EJE-16-0357. - DOI - PMC - PubMed
    1. Payankaulam S, Raicu AM, Arnosti DN. Transcriptional regulation of INSR, the insulin receptor gene. Genes. . 2019;10:984. doi: 10.3390/genes10120984. - DOI - PMC - PubMed

LinkOut - more resources