Factors affecting bone age maturation during 3 years of growth hormone treatment in patients with idiopathic growth hormone deficiency and idiopathic short stature: Analysis of data from the LG growth study.

To investigate the progression rate of bone age (BA) and associated factors during the first 3 years of growth hormone (GH) treatment in children with idiopathic GH deficiency (iGHD) and idiopathic short stature (ISS).Data for prepubertal children with iGHD and ISS who were treated with recombinant human GH were obtained from the LG Growth Study Database and analyzed. Height, weight, BA, insulin-like growth factor-1 (IGF-1) level, and GH dose were recorded every 6 months. Differences between BA and chronological age (CA), BA-CA, were calculated at each measurement. This study included 92 (78 iGHD and 14 ISS) subjects.After 3 years of GH treatment, the height z-score was -1.09 ± 0.71 (P < .001 compared to baseline), BA-CA was -1.21 ± 1.18 years (P < .001), and IGF-1 standard deviation score (SDS) was 0.43 ± 1.21 (P < .001) in the iGHD subjects; the change in BA over the 3 years was 3.68 ± 1.27 years. In the ISS subjects, the height z-score was -1.06 ± 0.59 (P < .001), BA-CA was -0.98 ± 1.23 years (P = .009), and IGF-1 SDS was 0.16 ± 0.76 (P = .648); the change in BA over the 3 years was 3.88 ± 1.36 years. The only significant factor associated with the BA progression was the BA-CA at 1 year of GH treatment (OR = 2.732, P = .001). The baseline BA-CA, IGF-1 SDS, and GH dose did not influence BA progression.Prepubertal subjects with iGHD and ISS showed height improvement and mild BA acceleration over the first 3 years of GH treatment. However, because the BA progression rate was considered to be clinically acceptable, GH treatment may increase the predicted adult height during this period.

Introduction

Many signals are involved in growth plate maturation, including growth hormone (GH), insulin-like growth factor-1 (IGF-1), glucocorticoids, thyroid hormone, estrogen, androgen, vitamin D, leptin, paracrine factors, extracellular matrix, and intracellular mechanisms.[ GH and IGF-1 are traditionally considered potent stimulators of bone growth.[ In children with GH deficiency (GHD), bone age (BA) is significantly delayed compared to chronological age (CA). After GH treatment, increased serum IGF-1 or GH levels stimulate growth plate development and result in BA progression. Usually, BA progression rates within 1 year of CA are considered normal, but some studies have found that various rates of skeletal maturation can be associated with underlying diseases,[ pubertal status, and obesity.[ BA changes can be erratic over time, even in normal healthy boys.[

The goal of GH treatment is to increase final adult height. When predicting adult height, the Bayley and Pinneau method use specific tables for delayed, average, and advanced BA which are determined by the Greulich and Pyle hand standards.[ Therefore, the BA progression rate during GH treatment has an important impact on adult height prediction. Nonetheless, the BA progression rate of those who receive GH treatment varies and can be either within the normal range[ or accelerated,[ and the results may be affected by age, sex, underlying disease, GH dose, duration of GH treatment, or pubertal status.

As the height at onset of puberty and final height are closely associated, maximizing the duration of prepubertal GH treatment may be beneficial.[ However, data on the BA progression rate of prepubertal patients with GHD over long-term GH treatment are insufficient. Therefore, we investigated the progression rate of BA during the first 3 years of GH treatment in prepubertal children with idiopathic GHD (iGHD) and idiopathic short stature (ISS) based on data obtained from the LG Growth Study (LGS). We also analyzed factors associated with the rapid progression of BA in children with iGHD.

Materials and methods

Data of prepubertal children with iGHD and ISS who were treated with recombinant human GH (Eutropin inj., EutropinAQ inj., and EutropinPlus inj., LG Chem, Ltd, Korea) were obtained from the LGS Database and assessed in our study. The LGS has been in progress since 2012 and is an open-label, multicenter, prospective, and retrospective observational study.[ Informed consent was obtained from all participants before the study enrollment period. The institutional review board of Hallym University Sacred Heart Hospital approved this study (IRB# 2017-I129).

Idiopathic GHD was defined when all of the following criteria were met: (1) height below the third percentile; (2) peak GH levels below 10 μg/L in two standard stimulation tests; and (3) BA delayed compared to CA.

Idiopathic short stature was defined when a child with normal birth weight and height was below the third percentile of height, had normal GH responses in stimulation tests and no chromosomal abnormality, and lacked any other identifiable disease related to short stature. The participants were treated with GH for at least 3 years, and the height, weight, body mass index (BMI), pubertal status, BA, IGF-1, and IGF-binding protein-3 (IGFBP-3) values were recorded at baseline and during follow-up. The GH dose was individualized after the recommended initial dose. The posology was then adapted at the first-, second-, and third-year follow-up evaluations. Only subjects in the prepubertal stage at baseline and during the follow-up period were included. BA was determined by the treating physician and was based on the standards of Greulich and Pyle.[ Differences between BA and CA were calculated at each measurement and defined as follows: BA0-CA0, at baseline; BA1-CA1, at the end of the first year of GH treatment; BA2-CA2, at the end of the second year of GH treatment; BA3-CA3, at the end of the third year of GH treatment. All anthropometric measurements were converted to z-scores using a Korean growth standard.[ The IGF-1 standard deviation score (SDS) and IGFBP-3 SDS values were calculated using the Korean normal reference.[ Subjects with at least 1 missing anthropometric or BA measurement during the 3-year follow-up period were excluded. Finally, 92 (78 iGHD and 14 ISS) subjects were included in this study.

Statistical analysis

All data are presented as the mean ± standard deviation (SD). Differences in anthropometric measurements and BA between the iGHD and ISS subjects were compared using Student t test or Wilcoxon rank sum test. Differences in the BA-CA values during GH treatment, compared to their basal values, were investigated by paired t tests. Serial changes of anthropometric measurements and BA were analyzed using a mixed-effects model repeated measures (MMRM) approach. Comparisons of BA progression according to baseline BA characteristics were analyzed using McNemar test. A comparison of factors associated with the rapid progression of BA at 3 years of GH treatment (BA3-CA3 ≥−1.00 year) was performed using multiple logistic regression analysis. All statistical analyses were performed using SAS software version 9.4 (SAS Institute, Cary, NC). P-values of <.05 were considered significant.

Results

Clinical characteristics of subjects (Table 1)

Of the 78 subjects (47 boys, 31 girls) with iGHD, the mean age at baseline (before GH treatment) was 7.80 ± 2.77 years. The height and BMI z-scores were −2.45 ± 0.67 and −0.24 ± 1.08, respectively. Their BA was 1.99 ± 0.96 years delayed compared to their CA. The IGF-1 SDS and IGFBP-3 SDS values were −0.89 ± 0.90 and −0.29 ± 1.76, respectively. After 3 years of GH treatment, the height z-score was −1.09 ± 0.71, which was significantly different from to the baseline height (P < .001), BA3-CA3 was −1.21 ± 1.18 years (P < .001 compared to BA0-CA0), and the BA change over the 3 years was 3.68 ± 1.27 years. The IGF-1 SDS and IGFBP-3 SDS values were 0.43 ± 1.21 (P < .001 compared to baseline IGF-1 SDS) and 0.67 ± 2.09 (P = .034 compared to baseline IGFBP-3 SDS), respectively, with a GH dose of 0.25 ± 0.12 mg/kg/wk. There were no differences between males and females in the clinical characteristics at baseline or during follow-up, except for BA3-CA3 (males; −1.57 ± 1.16 vs females; −0.66 ± 1.00; P = .001).

Of the 14 subjects (8 boys, 6 girls) with ISS, the mean age at baseline was 8.14 ± 2.97 years. The height and BMI z-scores were −2.60 ± 0.62 and −0.28 ± 0.93, respectively. Their BA was 2.04 ± 1.25 years delayed compared to their CA. Their IGF-1 SDS and IGFBP-3 SDS values were −0.31 ± 0.78 and −0.93 ± 0.88, respectively. After 3 years of GH treatment, the height z-score was −1.06 ± 0.59 (P < .001 compared to baseline height z-score). BA3-CA3 was −0.98 ± 1.23 years (P = .009 compared to BA0-CA0), and the BA change over the 3 years was 3.88 ± 1.36 years. The IGF-1 SDS and IGFBP-3 SDS values were 0.16 ± 0.76 (P = .648 compared to baseline IGF-1 SDS) and −0.62 ± 0.24 (P = .881 compared to baseline IGFBP-3 SDS), respectively, with a GH dose of 0.22 ± 0.09 mg/kg/wk.

The baseline characteristics of height, BMI, and BA-CA showed no significant differences between the iGHD and ISS cohorts. Height z-score, BMI z-score, BA, IGF-1 SDS, and IGFBP-3 SDS changes during the 3 years of GH treatment did not differ between the iGHD and ISS cohorts.

BA-CA and IGF-1 SDS changes during the 3 years of GH treatment (Figs. 1 and 2)

In subjects with iGHD, mean BA-CA values were −1.80 ± 1.03 years at the end of 1 year of GH treatment (P = .074 compared to BA0-CA0), −1.43 ± 0.97 years at the end of 2 years (P < .001 compared to BA0-CA0), and −1.21 ± 1.18 years at the end of 3 years (P < .001 compared to BA0-CA0). BA-CA changes were significant in the second year of GH treatment (P for trend = .024 by the MMRM approach, Fig. 2A). Changes in BA during the 1-year intervals were 1.21 ± 0.82 years (in the first year of GH treatment), 1.35 ± 0.86 years (in the second year of GH treatment), and 1.12 ± 0.75 years (in the third year of GH treatment). The IGF-1 SDS increased prominently during the first year of GH treatment (P < .001 compared to baseline IGF-1 SDS), but no significant changes were found thereafter (Fig. 2A).

Figure 2

IGF-1 SDS changes during GH treatment in subjects with iGHD (A) and ISS (B) are shown. ∗P < .05 compared to baseline IGF-1 SDS by the MMRM approach. GH = growth hormone, IGF-1 = insulin-like growth factor-1, iGHD = idiopathic GHD, ISS = idiopathic short stature, MMRM = mixed-effect model repeated measure, SDS = standard deviation score.

BA-CA changes during GH treatment in subjects with iGHD (A) and ISS (B) are shown. ∗P < .05 compared to baseline BA-CA by the MMRM approach. BA = bone age, CA = chronological age, GH = growth hormone, iGHD = idiopathic GHD, ISS = idiopathic short stature, MMRM = mixed-effect model repeated measure.

In subjects with ISS, mean BA-CA values were −1.67 ± 1.70 years at the end of 1 year of GH treatment (P = .129 compared to BA0-CA0), −1.23 ± 1.48 years at the end of 2 years (P = .007 compared to BA0-CA0), and −0.98 ± 1.23 years at the end of 3 years (P = .009 compared to BA0-CA0). BA-CA changes did not show significant differences during the 3 years of GH treatment (Fig. 2B). Changes in BA in 1-year intervals were 1.39 ± 0.84 years (in the first year of GH treatment), 1.42 ± 0.72 years (in the second year of GH treatment), and 1.06 ± 1.00 years (in the third year of GH treatment). The IGF-1 SDS increased prominently during the first year of GH treatment (P = .003 compared to baseline IGF-1 SDS), but the IGF-1 SDS then exhibited a decreasing trend (Fig. 2B).

Factors associated with BA progression during GH treatment in iGHD subjects

BA progression greater than 3 years during GH treatment was found in 82.1% (64/78) of the subjects with iGHD. The BA progressions showed significant differences according to delayed BA baseline status (P < .001, Fig. 3). Of the 65 subjects who showed a BA0-CA0 <−1.00 year, 63.1% had a BA3-CA3 <−1.00 year, while 69.2% of the 13 subjects who showed a BA0-CA0 ≥−1.00 year had a BA3-CA3 ≥−1.00 year. However, the only significant factor associated with BA3-CA3 was BA1-CA1 (OR = 2.732; 95% CI, 1.495–4.993; P = .001, Fig. 4), after adjusting for other variables, including sex, age, and baseline BA delay. BA0-CA0 showed positive relationships with BA3-CA3 in a simple correlation analysis (r = 0.333, P = .003), but these correlations were not significant in the multiple logistic regression analysis. The IGF-SDS and IGFBP-3 SDS at the end of the third year of GH treatment and the GH dose did not influence BA3-CA3.

Figure 3

Progression of bone age during GH treatment according to baseline characteristics in subjects with iGHD. GH = growth hormone, iGHD = idiopathic GHD.

Figure 4

BA1-CA1 was significantly positively associated with BA3-CA3 (OR = 2.732; 95% CI: 1.495–4.993; P = .001). BA1-CA1 = difference between BA and CA at the end of the first year of GH treatment, BA3-CA3 = difference between BA and CA at the end of the third year of GH treatment. BA = bone age, CA = chronological age, CI = confidence interval, GH = growth hormone, OR = odds ratio.

Discussion

In summary, 3 years of GH treatment resulted in significant height improvements in both iGHD and ISS cohorts, as based on height z-scores. The progression of BA (iGHD, 3.68 years; ISS 3.88 years) was slightly greater than that of CA during the 3 years of GH treatment; however, the mean values were considered clinically acceptable. The BA progression/year of CA was not accelerated over the treatment period (iGHD, 1.12 years and ISS, 1.06 years at the third year of GH treatment). Finally, the progression of BA after the first year of GH treatment affected BA maturation over the 3 years of GH treatment in subjects with iGHD, and contrary to our expectations, IGF-1 SDS changes were not relevant to BA progression.

The progression rate of BA during GH treatment in subjects with iGHD has been reported in previous studies to vary between 0.7 and 1.6 years/year of CA.[ Darendeliler et al. suggested that a BA progression of less than 2 yr/yr of CA can be considered normal, at least during the prepubertal period,[ and Benso et al found that BA progression varied from 0.5 yr/yr to over 2 yr/yr in normal boys.[ A study with similar conditions to ours (prepubertal at the start of GH treatment, a GH dose of 0.35 mg/kg/wk and 2 years of GH treatment) by Cohen et al reported 1.2 years of BA progression/year of CA,[ which was consistent with our findings (1.28 years of BA progression/year of CA). Many confounding factors may impact these diverse results, but among them, pubertal status is an important factor. Prolonged exposure to elevated levels of estradiol or testosterone is needed to markedly accelerate BA maturation during puberty.[ Kawai et al[ and Hopwood et al[ studied the efficacy of GH treatment in terms of improving final adult height in non-GHD or GHD subjects presenting with short stature and found 2 opposite results according to pubertal status; treatment was beneficial when the subject remained prepubertal versus not beneficial when the subject became pubertal during treatment. To control this important confounding factor, we only included prepubertal subjects. Our findings suggest that 3 years of GH treatment at a conventional dose in prepubertal subjects with iGHD did not adversely affect BA maturation.

ISS is a growth disorder that is associated with normal stimulation of GH secretion, and the degree of BA delay and the recommended initial GH dose for patients with ISS differ from those of iGHD.[ However, in our study, baseline clinical characteristics, including GH dose and IGF-1 SDS, showed no differences between the iGHD and ISS groups. This could lead to similar BA progression patterns in subjects with ISS and those with iGHD, and this observation has also been found in other studies.[

Regarding factors affecting BA maturation, previous studies have indicated that baseline BA delay is positively correlated with BA progression.[ Our study identified similar correlation patterns between the progression rate of BA and baseline BA delay; however, when other factors were considered, the statistical significance was marginal. Interestingly, BA delay at the end of the first year of GH treatment was found to be a significant factor that predicted BA delay at the end of the third year of GH treatment in our study. This result may be associated with IGF-1 SDS changes. Independent of sex or pubertal stage, IGF-1 plays a well-known role in skeletal maturation.[ In previous studies, increased IGF-1 levels due to GH treatment were found to be the main factor for growth response,[ and BA/CA groups, according to tertiles, showed positive correlations with IGF-1 z-scores among subjects who received GH treatment.[ In our study, IGF-1 SDS changes displayed a sharp increase during the first year of GH treatment, which could have significantly affected the BA1-CA1. However, the IGF-1 SDS did not increase after the first year, while height z-scores consistently improved over time during GH treatment, in both the iGHD and ISS cohorts. We did not identify a direct correlation between the IGF-1 SDS and BA progression; however, these findings emphasize the meaning of BA1-CA1 in this study.

There is an unavoidable difficulty in the method of BA estimation used in this study. Because this was a multicenter study, inter-observer differences occurred when determining BA. We used the interpretation of BA based on the Greulich and Pyle hand atlas, which could be subjective. However, when considering the prediction of adult height, the Greulich and Pyle method are suggested to be more accurate than other BA assessment methods such as the Tanner-Whitehouse 2 method.[ Additionally, the LGS data were valuable because they originated from a multicenter database, which could attenuate investigator bias.

In conclusion, both the iGHD and ISS prepubertal cohorts showed height improvement and mild BA acceleration over the first 3 years of GH treatment. However, because the BA progression rate is considered clinically acceptable, GH treatment may increase the predicted adult height during this period.

Acknowledgments

We thank all of the physicians who contributed their patient data to the LG Growth Study.

Author contributions

Conceptualization: Min Jae Kang, Il Tae Hwang.

Data curation: Min Jae Kang, Eun Young Kim.

Formal analysis: Min Jae Kang, Eun Young Kim.

Methodology: Min Jae Kang, Eun Young Kim.

Project administration: Min Jae Kang.

Resources: Eun Young Kim.

Software: Eun Young Kim.

Supervision: Seung Yang, Il Tae Hwang.

Validation: Young Suk Shim, Hwal Rim Jeong, Hye Jin Lee, Seung Yang.

Visualization: Young Suk Shim, Hwal Rim Jeong, Hye Jin Lee, Seung Yang.

Writing – original draft: Min Jae Kang.

Writing – review and editing: Min Jae Kang, Il Tae Hwang.

Min Jae Kang orcid: 0000-0003-3080-0941.

Il Tae Hwang orcid: 0000-0002-3885-4322.

Table 1

Clinical characteristics of subjects during GH treatment.