Impact of growth hormone preparation switching on treatment outcomes in children with growth hormone deficiency: a retrospective real-world cohort study

Abstract

Background

Switching between recombinant human growth hormone (rhGH) preparations has become increasingly common in pediatric practice because of reimbursement policies, product availability, and patient preferences. However, concerns persist regarding potential effects on growth outcomes and biochemical responses in real-world settings. This study aimed to evaluate whether switching growth hormone (GH) preparations is associated with differences in growth and biochemical outcomes compared with continuing the same product in children with growth hormone deficiency (GHD).

Methods

This single-center retrospective cohort study included 130 children with GHD receiving GH therapy between 2019 and 2024: 66 who underwent a single switch between preparations and 64 who continued the same product. Annualized changes in height standard deviation score (SDS), height velocity, insulin-like growth factor-1 (IGF-1) SDS, and insulin-like growth factor binding protein-3 (IGFBP-3) SDS were assessed over predefined 12-month observation windows. Patients were additionally stratified according to GHD severity, defined by a peak GH response < 7 ng/mL (severe/complete GHD) or 7–10 ng/mL (partial/borderline GHD). Within-group comparisons were performed using Wilcoxon signed-rank or Friedman tests, and between-group comparisons were performed using Mann–Whitney U tests or chi-square tests, as appropriate. Multivariate linear regression analyses adjusted for age, pubertal status, GH dose, baseline height SDS, and GHD severity.

Results

Baseline demographic, auxological, and perinatal characteristics were similar between the groups. In both cohorts, height SDS, IGF-1 SDS, and IGFBP-3 SDS improved significantly during follow-up, whereas GH dose remained stable. When outcomes were compared during equivalent treatment intervals, no significant differences were observed between the switch and nonswitch groups in height velocity (8.72 ± 2.64 vs. 8.42 ± 1.62 cm/year; p = 0.108), ΔHeight SDS (0.53 ± 0.51 vs. 0.39 ± 0.37; p = 0.206), ΔIGF-1 SDS (0.82 ± 0.69 vs. 0.71 ± 0.54; p = 0.486), or ΔIGFBP-3 SDS (0.86 ± 0.90 vs. 0.64 ± 0.45; p = 0.283). Similar findings were observed in patients with severe GHD (peak GH < 7 ng/mL). Multivariate regression analysis identified baseline height SDS and GH dose as independent predictors of 12-month height SDS change, whereas switch status and GHD severity were not independently associated with growth response.

Conclusions

Switching between rhGH preparations was not associated with impaired growth or adverse biochemical outcomes in children with GHD. Growth and biochemical responses were comparable between switched and nonswitched patients during equivalent treatment intervals, including among patients with severe GHD. These findings support the clinical feasibility of switching between daily rhGH preparations when required because of reimbursement policies, product availability, or other practical clinical considerations.

Trial registration

Not applicable.

Background

Recombinant human growth hormone (rhGH) therapy has been widely used since 1985 for children with growth hormone deficiency (GHD) and a variety of other growth-related disorders [1–4]. Its efficacy and safety have been well established in randomized trials and long-term observational cohorts [5–7]. The most pronounced catch-up growth typically occurs during the first two years of treatment, although individual responses vary considerably according to underlying diagnosis, age at treatment initiation, pubertal status, adherence, dosing strategies, and diagnostic criteria [8–11].

In contemporary clinical practice, switching between growth hormone (GH) preparations has become increasingly common owing to supply policies, reimbursement frameworks, the widespread introduction of biosimilar products, and patient or caregiver preferences [12, 13]. Although licensed rhGH formulations are generally regarded as clinically equivalent, subtle differences in injection devices, tolerability, or pharmacokinetic profiles may influence adherence and treatment perception [1, 2, 12, 13]. Consequently, clinicians and families frequently express concern regarding the potential impact of switching on growth outcomes.

The evidence addressing these concerns remains incomplete, particularly with respect to real-world pediatric cohorts with careful adjustment for pubertal timing and treatment duration [14–17]. Several registry-based analyses and real-world cohort studies have reported comparable auxological and biochemical responses after switching between originator and biosimilar GH preparations, supporting the clinical interchangeability of licensed products [18–20]. Nevertheless, real-world pediatric data from different healthcare systems remain relatively scarce, and evidence from carefully characterized cohorts with adjustment for pubertal timing and treatment duration is still limited [14–17].

Skeletal maturation and pubertal progression are additional determinants of the growth response during GH therapy, and bone age advancement does not always parallel linear growth [8, 21–23]. These factors underscore the importance of evaluating GH switching in carefully characterized cohorts with longitudinal auxological and biochemical follow-up [18–20].

Therefore, the present study aimed to examine, in a real-world pediatric cohort of children with GHD, whether switching between rhGH preparations is associated with differences in growth response or biochemical markers compared with continuing the same product. We focused on annualized changes in the height standard deviation score and growth velocity while accounting for key clinical covariates, including age, pubertal status, GH dose, treatment duration, and skeletal maturation [10, 15, 24].

Methods

Study design and population

This single-center retrospective cohort study was conducted at the Department of Pediatric Endocrinology between September 2019 and March 2024. Children receiving recombinant human growth hormone (rhGH) therapy with Norditropin, Saizen, Genotropin, or Omnitrope for growth hormone deficiency (GHD) were eligible for inclusion. All evaluated GH preparations were daily rhGH formulations administered once daily throughout the study period.

The diagnosis of GHD was established according to standard criteria [3, 5]. Patients were evaluated after at least one year of clinical follow-up. Eligibility for GH stimulation testing required a height standard deviation score (SDS) below − 2, an age-inappropriate height velocity, delayed bone age, and low serum insulin-like growth factor-1 (IGF-1) and/or insulin-like growth factor-binding protein-3 (IGFBP-3) levels. GH stimulation tests were performed using clonidine and L-dopa, and a peak GH response below 10 ng/mL on both tests was required to confirm the diagnosis of GHD. For subgroup analyses, patients with a peak GH response < 7 ng/mL were classified as having severe/complete GHD, whereas those with peak GH responses between 7 and 10 ng/mL were classified as having partial/borderline GHD.

Switching between GH preparations occurred because of national procurement and availability policies or clinical considerations such as device preference or convenience rather than inadequate treatment response. All patients in both groups were newly initiated on GH therapy at treatment commencement. In the switch group (n = 66), patients underwent a single documented switch to an alternative GH preparation after 12 months of treatment with the initial product, followed by at least 12 additional months of continued therapy with the new preparation. In the nonswitch group (n = 64), patients remained on the same GH preparation throughout the entire 24-month observation period without switching. The study was not designed to compare the efficacy of individual GH brands or manufacturers.

Patients with incomplete auxological or biochemical data, multiple switches, or insufficient follow-up data were excluded.

Clinical and biochemical assessments

Demographic characteristics, medical history, anthropometric measurements, pubertal stage, serum IGF-1 and IGFBP-3 concentrations (expressed as standard deviation scores [SDS]), GH dose, and bone age were extracted from medical records. Bone age was assessed annually via the Greulich–Pyle method [21].

GH was administered once daily at bedtime. The prescribed dose (µg/kg/day) was calculated on the basis of actual body weight and adjusted according to the growth response and serum IGF-1 concentration. Height and weight SDS values were derived from national Turkish reference standards [24].

Standing height was measured via a wall-mounted stadiometer with patients barefoot and positioned in the Frankfurt plane. All height measurements were performed by trained personnel using the same device according to standardized protocols. Weight was measured with a calibrated digital scale. To minimize interobserver variability, pubertal staging was assessed by a pediatric endocrinologist at each visit according to the Tanner criteria, and bone age evaluations were performed by the same pediatric endocrinologist via the Greulich–Pyle atlas [21].

Patients were routinely followed at 6-month intervals. Height velocity, GH dose, and biochemical parameters were recorded at each visit.

Definition of outcomes and exposure

Changes in height SDS (ΔHeight SDS), IGF-1 SDS (ΔIGF-1 SDS), and IGFBP-3 SDS (ΔIGFBP-3 SDS) were calculated as the difference between values obtained at the beginning and end of predefined 12-month observation windows.

In the switch group, the observation period spanned from 12 months before the switch (baseline) to 12 months after the switch, with the switch time point serving as the primary reference. Outcomes were separately evaluated for the 12-month preswitch and 12-month postswitch periods to enable within-group longitudinal comparisons. For between-group comparisons, outcomes during the 12-month postswitch period in the switch group were compared with outcomes during the second treatment year (months 12–24) in the nonswitch group.

Height velocity represented the average annualized growth rate during the corresponding 12-month observation period and was expressed in centimeters per year.

Statistical analysis

Statistical analyses were performed using SPSS for Windows version 22.0 (IBM Corp., Armonk, NY, USA). Categorical variables are presented as frequencies and percentages, whereas continuous variables are expressed as mean ± standard deviation (minimum–maximum). Pearson’s chi-square test was used for comparisons of categorical variables. For continuous variables that did not conform to a normal distribution, the Mann–Whitney U test was applied. The Wilcoxon signed-rank test was used to compare paired measurements, and the Friedman test was employed for comparisons of three or more related measurements.

To address potential heterogeneity in diagnostic severity, additional subgroup analyses were performed according to growth hormone deficiency severity. Patients with a peak GH response < 7 ng/mL were classified as having severe/complete GHD, whereas those with peak GH responses between 7 and 10 ng/mL were classified as having partial/borderline GHD.

Multivariate linear regression analysis was conducted to identify independent factors associated with 12-month height SDS changes, with switching status, age at treatment initiation, baseline height SDS, GH dose, baseline pubertal status, and GHD severity category entered as covariates. A p value < 0.05 was considered statistically significant.

Results

Baseline characteristics

According to Table 1, of the 130 patients with GH deficiency enrolled, 66 were assigned to the switch group and 64 to the nonswitch group. Baseline demographic characteristics were comparable between the groups, with no significant differences in sex distribution (33/33 vs. 40/24; p = 0.208), pubertal status (prepubertal/pubertal: 34/32 vs. 31/33; p = 0.861), or age (10.67 ± 2.52 vs. 9.69 ± 3.58 years; p = 0.074). Auxological parameters at baseline were also similar, including height SDS (− 2.49 ± 0.49 vs. −2.53 ± 0.77; p = 0.725) and BMI SDS (− 0.15 ± 1.06 vs. 0.13 ± 0.86; p = 0.100). Furthermore, target height (161.92 ± 8.94 vs. 162.15 ± 7.99 cm; p = 0.877), gestational age (37.92 ± 2.15 vs. 38.22 ± 1.96 weeks; p = 0.407), and birth weight (3060 ± 546 vs. 3061 ± 546 g; p = 0.992) did not differ significantly between the groups. The distribution of GHD severity was also comparable between groups, with severe GHD present in 35/66 (53.0%) patients in the switch group and 36/64 (56.3%) patients in the nonswitch group (p = 0.712).

Table 1.

Baseline characteristics of patients with GH deficiency in the switch and nonswitch groups

Parameter	Switch group (n = 66)	Nonswitch group (n = 64)	p value
Male/Female, n (%)	33 (50.0)/33 (50.0)	40 (62.5)/24 (37.5)	0.208
Prepubertal/Pubertal, n (%)	34 (51.5)/32 (48.5)	31 (48.4)/33 (51.6)	0.861
Age (years), mean ± SD (range)	10.67 ± 2.52 (2.0–15.0)	9.69 ± 3.58 (1.0–14.8)	0.074
Height SDS	−2.49 ± 0.49	−2.53 ± 0.77	0.725
BMI SDS	−0.15 ± 1.06	0.13 ± 0.86	0.100
Target height (cm)	161.92 ± 8.94	162.15 ± 7.99	0.877
Gestational age (weeks)	37.92 ± 2.15	38.22 ± 1.96	0.407
Birth weight (g)	3060 ± 546	3061 ± 546	0.992
Severe GHD (< 7 ng/mL), n (%)	35 (53.0)	36 (56.3)	0.712

SDS, standard deviation score; BMI, body mass index. Data are presented as means ± SD unless otherwise stated. Pearson chi-square test for categorical variables; Mann–Whitney U test for continuous variables

In the switch group, height SDS increased significantly from baseline (− 2.58 ± 0.46) to + 12 months (− 1.75 ± 0.90) over the follow-up period (p < 0.001). Height velocity was highest at 6 months before the switch (10.12 ± 3.03 cm/year) and showed a progressive decline thereafter, yet the within-group change remained significant (p < 0.001). IGF-1 SDS and IGFBP-3 SDS both improved significantly over time (p < 0.001), while GH dose did not undergo a significant change (p = 0.370). In the nonswitch group, height SDS also increased significantly from baseline (− 2.53 ± 0.77) to 24 months (− 1.64 ± 0.59) (p < 0.001), and height velocity declined from 12 months to 24 months (p < 0.001). IGF-1 SDS and IGFBP-3 SDS rose markedly over the follow-up period (p < 0.001), whereas GH dose remained unchanged (p = 0.423) (Table 2).

Table 2.

Within-group auxological and biochemical changes in the switch and nonswitch cohorts over the follow-up period

Switch Cohort (n = 66)
Parameter	12 months before switch	6 months before switch	Switch	6 months after switch	12 months after switch	p value†
Height SDS	−2.58 ± 0.46	−2.21 ± 0.65	−1.99 ± 0.78	−1.87 ± 0.82	−1.75 ± 0.90	< 0.001
IGF-1 SDS	−1.96 ± 0.49	−1.19 ± 0.72	−1.03 ± 0.91	−0.89 ± 0.75	−1.12 ± 0.87	< 0.001
IGFBP-3 SDS	−0.74 ± 0.82	0.06 ± 0.73	0.26 ± 0.73	0.27 ± 0.52	0.28 ± 0.72	< 0.001
Height velocity (cm/year)	—	10.12 ± 3.03	7.81 ± 2.57	6.92 ± 2.06	6.69 ± 2.65	< 0.001
Bone age (years)	8.65 ± 2.67	—	9.82 ± 2.44	—	11.31 ± 2.34	< 0.001
GH dose (µg/kg/day)	29.83 ± 3.90	28.59 ± 3.89	30.48 ± 3.84	30.83 ± 4.67	31.00 ± 4.76	0.370

Nonswitch Cohort (n = 64)
Parameter	Baseline	12 months	24 months	p value†
Height SDS	−2.53 ± 0.77	−2.04 ± 0.62	−1.64 ± 0.59	< 0.001
IGF-1 SDS	−1.86 ± 0.44	−0.92 ± 0.79	−0.80 ± 0.82	< 0.001
IGFBP-3 SDS	−0.64 ± 0.62	0.31 ± 0.81	0.35 ± 0.78	< 0.001
Height velocity (cm/year)	—	9.34 ± 2.35	7.51 ± 1.94	< 0.001
Bone age (years)	8.07 ± 3.62	9.28 ± 3.72	10.52 ± 3.79	< 0.001
GH dose (µg/kg/day)	29.02 ± 4.78	29.27 ± 3.87	29.70 ± 5.12	0.423

SDS, standard deviation score; IGF-1, insulin-like growth factor-1; IGFBP-3, insulin-like growth factor-binding protein-3; GH, growth hormone. The data are presented as the means ± SDs. p values refer to within-group longitudinal comparisons. *; Friedman analysis

Switch cohort: preswitch versus postswitch 12-month outcomes

When parameters were compared before and after switching within the GHD group, height velocity was found to decrease significantly after the switch (9.13 ± 2.49 cm/year vs. 7.28 ± 2.29 cm/year; p < 0.001). Similarly, significant reductions were observed in both ΔIGF-1 SDS (1.08 ± 0.84 vs. 0.26 ± 0.66; p < 0.001) and ΔIGFBP-3 SDS (0.90 ± 0.93 vs. 0.15 ± 0.71; p < 0.001) following the switch.In contrast, no significant difference was detected in ΔHeight SDS between the pre-switch and post-switch periods (0.48 ± 0.46 vs. 0.36 ± 0.79; p = 0.311). Although there was a trend toward an increased GH dose after switching, this change did not reach statistical significance 29.83 ± 3.90 vs. 31.00 ± 4.76 µg/kg/day; p = 0.151) (Table 3).

Table 3.

Within-group auxological and biochemical changes in the switch cohort before and after switching

Parameter	Preswitch	Postswitch	p value*
Height velocity (cm/year)	9.13 ± 2.49	7.28 ± 2.29	< 0.001
ΔHeight SDS	0.48 ± 0.46	0.36 ± 0.79	0.311
ΔIGF-1 SDS	1.08 ± 0.84	0.26 ± 0.66	< 0.001
ΔIGFBP-3 SDS	0.90 ± 0.93	0.15 ± 0.71	< 0.001
GH dose (µg/kg/day)	29.83 ± 3.90	31.00 ± 4.76	0.151

SDS, standard deviation score; IGF-1, insulin-like growth factor-1; IGFBP-3, insulin-like growth factor-binding protein-3; GH, growth hormone. The data are presented as the means ± SDs. p values indicate paired within-group comparisons. *; Wilcoxon signed-rank tests

Between-group comparison of annualized outcomes

As shown in Table 4, comparison of outcomes during comparable observation periods demonstrated no significant differences between the switch and nonswitch groups in height velocity (8.72 ± 2.64 vs. 8.42 ± 1.62 cm/year; p = 0.108), ΔHeight SDS (0.53 ± 0.51 vs. 0.39 ± 0.37; p = 0.206), ΔIGF-1 SDS (0.82 ± 0.69 vs. 0.71 ± 0.54; p = 0.486), or ΔIGFBP-3 SDS (0.86 ± 0.90 vs. 0.64 ± 0.45; p = 0.283). Similarly, among patients with severe GHD (peak GH < 7 ng/mL), no significant differences were observed between the switch and nonswitch groups regarding height velocity (7.59 ± 2.81 vs. 7.35 ± 1.35 cm/year; p = 0.136), ΔHeight SDS (0.58 ± 0.56 vs. 0.40 ± 0.38; p = 0.240), ΔIGF-1 SDS (0.86 ± 0.79 vs. 0.70 ± 0.48; p = 0.189), or ΔIGFBP-3 SDS (0.91 ± 0.93 vs. 0.86 ± 0.62; p = 0.274) (Table 4).

Table 4.

Between-group comparisons of growth and biochemical outcomes during comparable observation periods in the overall cohort and severe GHD subgroup

Parameter	Overall Cohort			Severe GHD Subgroup (Peak GH < 7 ng/mL)
	Switch group (n = 66)	Nonswitch group (n = 64)	p value	Switch group (n = 35)	Nonswitch group (n = 36)	p value
Height velocity (cm/year)	8.72 ± 2.64	8.42 ± 1.62	0.108	7.59 ± 2.81	7.35 ± 1.35	0.136
ΔHeight SDS	0.53 ± 0.51	0.39 ± 0.37	0.206	0.58 ± 0.56	0.40 ± 0.38	0.240
ΔIGF-1 SDS	0.82 ± 0.69	0.71 ± 0.54	0.486	0.86 ± 0.79	0.70 ± 0.48	0.189
ΔIGFBP-3 SDS	0.86 ± 0.90	0.64 ± 0.45	0.283	0.91 ± 0.93	0.86 ± 0.62	0.274

SDS, standard deviation score; IGF-1, insulin-like growth factor-1; IGFBP-3, insulin-like growth factor-binding protein-3; GHD, growth hormone deficiency. For the overall cohort, outcomes during the 12-month postswitch period in the switch group were compared with outcomes during the second treatment year (months 12–24) in the nonswitch group. The severe GHD subgroup analysis was restricted to patients with a peak GH response < 7 ng/mL. Data are presented as mean ± SD. p values indicate between-group comparisons using the Mann–Whitney U test

Multivariate regression analysis

Multivariate linear regression analysis demonstrated that baseline height SDS (B = − 0.196, 95% CI − 0.295 to − 0.097, p < 0.001) and GH dose (B = − 0.019, 95% CI − 0.037 to 0.000, p = 0.045) were independently associated with 12-month changes in height SDS. Switch status was not independently associated with height SDS change (B = 0.338, 95% CI − 0.186 to 0.490, p = 0.096). GHD severity (partial/severe), age at treatment initiation, and baseline pubertal status were also not independently associated with growth response. The overall model was significant (F = 7.332, p < 0.001) and explained 33.8% of the variance in height SDS change (R² = 0.338; adjusted R² = 0.292). No evidence of multicollinearity was observed among the independent variables (all VIF values < 2.1) (Table 5).

Table 5.

Multivariate linear regression analysis of factors associated with 12-month changes in height SDS

Variable	B	p value	95% CI
Intercept	0.633	0.061	-0.031 to 1.297
Partial/Severe GHD	0.113	0.102	-0.023 to 0.250
Age at treatment initiation	-0.020	0.156	-0.048 to 0.008
Baseline height SDS	-0.196	< 0.001	-0.295 to -0.097
Switch status	0.338	0.096	-0.186 to 0.490
GH dose	-0.019	0.045	-0.037 to 0.000
Baseline pubertal status	0.005	0.960	-0.183 to 0.192

GHD, growth hormone deficiency; GH, growth hormone; CI, confidence interval. B represents unstandardized regression coefficients. Model statistics: R² = 0.338, adjusted R² = 0.292, F = 7.332, p < 0.001

Discussion

In this retrospective real-world cohort of 130 children with growth hormone deficiency, switching between recombinant human GH preparations was not associated with impaired growth or adverse biochemical outcomes. Height SDS improved significantly in both groups, GH dose remained within therapeutic ranges, and favorable auxological and biochemical responses were observed throughout follow-up. Importantly, when outcomes were compared during equivalent treatment intervals, no significant differences were observed between the switch and nonswitch groups in height velocity, ΔHeight SDS, ΔIGF-1 SDS, or ΔIGFBP-3 SDS. Similar findings were observed among patients with severe GHD, supporting the robustness of the primary analysis. Overall, our findings support the clinical feasibility and safety of switching between rhGH preparations when clinically necessary.

rhGH therapy has been used in clinical practice for more than three decades and remains the cornerstone treatment for children with GHD. The increasing availability of biosimilar products and evolving procurement policies have made switching between GH preparations a common occurrence in routine clinical practice, raising questions among clinicians and families regarding potential effects on efficacy and safety [3].

Baseline demographic and auxological characteristics were comparable between the switch and nonswitch groups and closely resembled those of previously reported Turkish and international GHD cohorts [8, 9, 14–17]. A substantial proportion of patients initiated GH therapy during puberty, reflecting delayed referral patterns and real-life clinical scenarios reported in earlier national studies [15, 16]. These baseline similarities strengthen the internal validity of subsequent between-group comparisons.

With respect to growth response, no significant differences were observed between the switch and nonswitch groups in height velocity or ΔHeight SDS when outcomes were compared during equivalent treatment intervals. Similar findings were observed in the severe GHD subgroup, indicating that switching did not adversely affect short-term growth outcomes regardless of disease severity. Height SDS improved progressively in both cohorts throughout follow-up, underscoring the overall effectiveness of GH therapy irrespective of formulation changes. These findings suggest that switching between daily rhGH preparations does not compromise growth response and support the clinical interchangeability of licensed rhGH products in routine pediatric endocrine practice [10, 11, 25].

Serum IGF-1 and IGFBP-3 concentrations provide important insight into treatment adherence and metabolic safety, although their trajectories do not invariably parallel the linear growth response [5]. In our cohort, annualized increases in IGF-1 SDS and IGFBP-3 SDS were greater during the preswitch period than during the postswitch period, consistent with the characteristic endocrine response following GH therapy initiation and subsequent stabilization as steady-state adaptation occurs. Importantly, a comparable attenuation of biochemical response was observed in the nonswitch cohort, in which IGF-1 SDS plateaued between 12 and 24 months of treatment (− 0.92 vs. −0.80, respectively), consistent with the expected stabilization of the GH–IGF-1 axis during continued therapy. This parallel trajectory in both cohorts strongly suggests that the biochemical changes observed after switching reflect the natural time course of GH treatment response rather than a formulation-specific effect. This pattern has been well described in longitudinal registries and should be considered when interpreting biochemical trends across different treatment phases [10, 25].

GH dosing patterns remained broadly comparable between groups and fell well within internationally accepted therapeutic ranges [26], reducing the likelihood that major dose disparities alone accounted for observed auxological differences. Notably, multivariate regression identified baseline height SDS and GH dose as independent predictors of 12-month height SDS change, whereas switch status was not independently associated with growth response. Previous predictive models of GH responsiveness have consistently emphasized the importance of treatment dose, baseline auxological severity, and treatment adherence as major determinants of growth outcomes, often outweighing formulation-specific factors alone [10, 11, 25]. Our findings are broadly consistent with these established principles, suggesting that while GH preparation switching appears clinically safe, individualized therapeutic management remains the primary driver of successful long-term growth response. Skeletal maturation is inherently multifactorial and reflects the combined influences of chronological age, pubertal status, intrinsic GH sensitivity, genetic background, and treatment adherence [6, 7, 27]. Accordingly, careful titration of the GH dose and serial bone age monitoring remain central components of long-term management.

Within the switch cohort, bone age advanced progressively throughout the follow-up period while height velocity declined after the formulation change. This temporal pattern likely reflects the well-recognized physiological attenuation of growth acceleration after the first year of therapy rather than a direct consequence of switching [6]. Although accelerated skeletal maturation could theoretically constrain subsequent linear growth by promoting earlier epiphyseal fusion, bone age progression remained within the expected range in both groups [23, 28, 29]. Notably, bone age advancement was observed in both cohorts across the follow-up period, consistent with the known stimulatory effects of GH therapy on skeletal maturation, and did not appear to differ in a clinically meaningful way between the groups. These observations highlight the importance of integrated interpretation of auxological and skeletal data throughout long-term GH therapy.

Most prior studies examining the interchangeability of GH products have reported neutral effects on growth and biochemical outcomes. Large European registries and Swedish real-world cohorts reported no clinically meaningful loss of efficacy or safety after switching preparations [18, 20]. Comparable conclusions were drawn in a multicenter cohort of pediatric patients with GHD, in which switched patients demonstrated auxological and biochemical responses equivalent to those continuing the same product [19]. Collectively, these reports support regulatory assessments that licensed GH products are clinically interchangeable for most patients.

Consistent with these earlier reports, our findings support the clinical feasibility of GH preparation switching in routine practice. Following reanalysis using equivalent treatment intervals, no significant differences were observed between the switch and nonswitch groups in auxological or biochemical outcomes, further supporting the interchangeability of licensed daily rhGH preparations. Consistent with the between-group comparisons, switch status was not independently associated with 12-month height SDS change in multivariable regression analysis, further supporting the conclusion that switching between rhGH preparations does not materially affect growth outcomes. Importantly, GHD severity was not independently associated with growth response, and subgroup analyses restricted to patients with severe GHD yielded findings consistent with those observed in the overall cohort.

Although the biological mechanisms underlying potential differences in treatment response among GH formulations remain incompletely understood, chronic GH exposure may theoretically induce adaptive changes at the receptor or signaling level through alterations in JAK2–STAT5 pathways or suppressor of cytokine signaling (SOCS) activation [30, 31]. Similarly, modest pharmacokinetic variability among licensed formulations may influence tissue-level GH exposure despite regulatory bioequivalence [32, 33]. However, such mechanisms remain speculative, and our findings do not suggest clinically meaningful differences in growth or biochemical outcomes between formulations. Prospective studies incorporating pharmacokinetic profiling and standardized pubertal stratification may further clarify these issues.

From a clinical perspective, our findings provide reassurance that switching between rhGH preparations does not compromise growth efficacy or biochemical outcomes in children with GHD. Comparable results were observed in both the overall cohort and the severe GHD subgroup. Individualized dose optimization, adherence support, and longitudinal monitoring remain the principal determinants of successful GH therapy.

In summary, this real-world cohort demonstrates that switching between daily recombinant human GH preparations is not associated with impaired growth or adverse biochemical outcomes in children with GHD. Comparable auxological and biochemical responses were observed in both switched and nonswitched patients during equivalent treatment intervals, and these findings remained consistent among patients with severe GHD. These results complement existing registry data and support the clinical feasibility of switching between licensed daily rhGH preparations when clinically indicated. Prospective multicenter studies with longer follow-up and mechanistic endpoints are warranted to further characterize long-term growth regulation during GH therapy and to determine whether formulation changes exert any independent biological effects.

Strengths and limitations

To our knowledge, this study represents the first Turkish real-world cohort specifically designed to evaluate the clinical impact of switching between recombinant human growth hormone (rhGH) preparations in children with growth hormone deficiency (GHD). Major strengths include the relatively large sample size for a single-center pediatric endocrine study, the use of standardized diagnostic criteria for GHD, and the availability of detailed longitudinal auxological, biochemical, and skeletal maturity data collected during routine clinical care. In addition, the inclusion of subgroup analyses according to GHD severity and multivariable adjustment for potential confounders strengthened the robustness of the findings. The standardized assessment of height, pubertal stage, and bone age by pediatric endocrinologists further enhanced internal consistency and reduced measurement variability.

Nevertheless, several limitations warrant consideration. First, the retrospective observational design precludes causal inference and leaves room for residual confounding despite multivariable adjustment. Although baseline demographic and clinical characteristics were comparable between the switch and nonswitch groups, the study was not designed as a matched cohort analysis. Therefore, residual confounding related to patient selection cannot be completely excluded. Future prospective studies incorporating individual matching or propensity score–based approaches may provide additional methodological rigor and further reduce the risk of selection bias. Second, objective adherence data were not available, and IGF-1 SDS was used as an indirect proxy for treatment compliance. Third, switching often involves changes in injection devices that may influence treatment experience and adherence; however, device-related information was not systematically recorded. Finally, postswitch follow-up was limited to 12 months, restricting evaluation of longer-term outcomes such as near-adult height and final adult height attainment.

Conclusions

In this real-world cohort of children with GHD, switching between daily recombinant human GH preparations was not associated with impaired growth or adverse biochemical outcomes. Comparable auxological and biochemical responses were observed in switched and nonswitched patients during equivalent treatment intervals, including among patients with severe GHD. These findings support the clinical feasibility and safety of switching between licensed daily rhGH preparations when clinically indicated.

When required because of reimbursement policies, product availability, or practical considerations, switching GH preparations can be undertaken without compromising short-term treatment outcomes. Nevertheless, individualized longitudinal monitoring remains essential, including assessment of growth velocity, skeletal maturation, pubertal progression, GH dose adjustment, and serum IGF-1 concentrations throughout treatment.

Prospective multicenter studies with longer follow-up to near-adult height and final adult height, as well as mechanistic investigations, are warranted to further characterize long-term growth outcomes and determine whether formulation changes exert any independent biological effects beyond expected therapeutic variability.

Abbreviations

BMI

Body mass index

GH

Growth hormone

GHD

Growth hormone deficiency

IGF

1–Insulin–like growth factor–1

IGFBP

3–Insulin–like growth factor binding protein–3

rhGH

Recombinant human growth hormone

SDS

Standard deviation score

SOCS

Suppressor of cytokine signalling

Author contributions

D.D., İ.G., and C.A.G. contributed to data collection and analysis. D.D., E.I., and P.K. drafted the manuscript. G.B., H.N.S., S.K.C., D.T., M.B., and F.G. critically reviewed and revised earlier versions of the manuscript. All authors contributed to the study conception and design, approved the final manuscript, and agreed to be accountable for all aspects of the work.

Funding

The authors received no specific funding for this work.

Data availability

The datasets generated and/or analysed during the current study are available from the corresponding author upon reasonable request and subject to institutional approval and ethical regulations.

Declarations

Ethics approval and consent to participate

This retrospective study involving human participant data was approved by the Institutional Ethics Committee of Ankara Bilkent City Hospital (Approval No: 2-24-516; Date: 18 September 2024). The requirement for informed consent to participate was waived by the ethics committee due to the retrospective design of the study, which used anonymized electronic medical records without direct patient contact. All procedures were conducted in accordance with the ethical standards of the institutional research committee and with the Declaration of Helsinki.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.