Sex steroid priming for growth hormone stimulation testing in children and adolescents with short stature: A systematic review

Abstract Objective Growth hormone stimulation testing (GHST) is used to diagnose growth hormone deficiency (GHD) in children. As sex steroids impact on anterior pituitary function, there is concern around the efficacy of GHST in peripubertal children, where endogenous sex steroid levels are low. Sex steroid priming before GHST is thought to improve test efficacy in these children, however evidence to support its use in clinical practice is limited. In this systematic review, we addressed the following research questions: Does priming increase GH stimulation test efficacy in peripubertal children? Does priming identify those who would benefit most from treatment in terms of final height? Is there evidence for an optimal sex‐steroid priming regimen? Design, Patients, Measurements The study was registered with PROSPERO and conducted according to PRISMA guidelines. We searched Medline, Cochrane‐Library, Scopus, EMBASE and Web‐of‐Science and included all studies that included GHST in both primed and unprimed children. A GH cut‐off of 7 µg/L was used as a threshold for GHD. Study quality was assessed using the Risk‐Of‐Bias in Non‐ Randomized Studies (ROBINS‐I) tool or the revised Cochrane risk‐of‐bias tool for Randomised trials. Results Fifteen studies met our inclusion criteria, of which 4/15 (27%) were randomised control trials. The majority (9/15) of the studies indicated that priming increases growth hormone response upon GHST in peripubertal children, increasing test specificity. Two studies investigated final height after treatment based on the results of primed versus unprimed GHST. These results indicate that growth hormone treatment based on results of a primed GHST improve outcomes compared with treatment based on an unprimed test. Conclusion Sex‐steroid priming increases the growth hormone response during GHST, resulting in fewer patients meeting the threshold required for a diagnosis of GHD. Unnecessary GH treatment may be avoided in some patients without a detrimental effect on final height. Numerous sex‐steroid priming regimens have been used in clinical practice and the majority appear to be effective, but an optimal regimen has not been determined.

been used in clinical practice and the majority appear to be effective, but an optimal regimen has not been determined. Sex steroids may modulate GH secretion and have the potential to impact the results of GHST. 5 In males, testosterone has been shown to stimulate growth hormone secretion, at least in part due to its aromatization to oestradiol. 6 In females, the effect of oestrogen on growth hormone production is less clear. It may have a direct effect on the pituitary stimulating GH production (although mechanisms of this remains uncertain) and indirect effects on GH production by reducing hepatic IGF-1 production resulting in secondary increase in GH secretion. 7,8 It is therefore important to consider that in the peripubertal agerange, prepubertal patients with low levels of circulating sex steroids may have a blunted response to GHST than pubertal age-matched peers. In these children, priming, which involves administering exogenous sex steroids before testing to sensitise the pituitary gland and potentiate GH secretion during GHST, can be performed. Even in children and adolescents with normal stature or idiopathic short stature, GH responses to various stimuli can be inconsistent and frequently do not reach the conventional cut-offs of 7 or 10 μg/L. 9 In addition, unprimed GHST have been reported to have poor specificity for diagnosing GHD in this cohort, while primed tests have been shown to reduce false-positive diagnoses of GHD. [10][11][12][13] The use of priming is still a matter of debate. The potential benefits that priming increases GHST specificity, reducing the rate of false positive results must be weighed against the concept that priming can affect physiology, producing only a transient rise in GH secretion, and that children who might benefit from GH therapy could be missed. 14,15 Guidelines on the diagnosis and management of GHD in children and adolescents have placed emphasis on avoiding unnecessary GH treatment in healthy patients with ISS and/or CDGP, considering its potential harms, costs and physical/ psychological burden. 16,17 Guidelines specifically recommending priming in short (adult height [AH] prognosis<-2 SDS) prepubertal boys aged over 11 and girls aged over 10, aiming to prevent unnecessary r-GH treatment in those where CDGP is likely have been produced, 16 whilst others were not able to reach consensus on the use of sex steroid priming in GHST. 17 It has also been suggested that priming may improve GHST specificity in prepubertal children of any age. 11,12,18 In addition to the issue of specificity, a recent audit has shown that in clinical practice, the regimens used for priming vary significantly. 19 Despite the importance of accurate diagnosis and treatment of GHD a systematic review of the evidence for sex-steroid priming for GHST has not previously been conducted. Therefore, the aim of our study was to perform a systematic review and meta-analysis of the literature, focusing on the impact of sex-steroid priming on the results of GHST and the influence of treatment based on a 'primed' or 'unprimed' test on final height.

| METHODS
We performed a comprehensive search of published clinical studies involving sex-steroid priming for GHST. This review is reported according to the PRISMA guidelines for systematic reviews. 20 The study protocol was registered as a systematic review on Prospero (registration number CRD42021244443) and is available from: https://www.crd.york.ac.uk/prospero/display_ record.php?ID=CRD42021244443.

| Information source
The search was included the following databases: Medline, Cochrane Library, Scopus, EMBASE and Web of Science.

| Inclusion criteria
Inclusion criteria for study selection were: • Articles published between January 1, 1900 and November 20, 2020 and written in English; • Participants were Children (0-9 years) and/or adolescents (10-19 years); • Includes 'primed' and 'unprimed' patients • Outcomes include effect of sex-steroid priming on outcome of GHST or final adult height

| Exclusion criteria
Exclusion criteria for the study selection were: • Studies not in the English language • Conference abstracts and reviews

| Search and study selection
The search strategy and terms (Supporting Information: Table 1) were adapted from an earlier study. The number of records identified after removal of duplicates was 66; initial screening of titles and abstracts was performed independently by two assessors (GD and RTM) using selection based on inclusion/exclusion criteria. For abstracts that were not selected by both assessors, a final decision on inclusion was made following discussion between assessors. Full texts were obtained for 32 studies and a further 18 articles were excluded as a result of not meeting all of the inclusion criteria (Supporting Information: Table 2). An additional article was identified on screening the reference lists of included papers. Fifteen papers were included in the final analysis. A summary of the selection process is provided in the PRISMA flowchart ( Figure 1).

| Data extraction and summary measures
The following data-(1) author (2) year of publication (3) study design (4) cohort size (5) sex steroid used (6) GHST used-were extracted separately by GD and RTM. Further data were extracted for data analysis: (7) population characteristics (8) pretest criteria (inclusion criteria for subjects) (9) control group included (10) study results. Additionally, each study underwent a risk of bias assessment. Risk of bias was assessed using the revised Cochrane risk-of-bias tool for Randomised trials (Rob 2) (Supporting Information:

| Statistics
Data for individual studies was represented according to reported mean ± 95% confidence interval and statistical significance was p < .05. Data for composite was presented as mean ± SEM and statistical analysis was performed using a Wilcoxon matched-pairs signed rank test, with statistical significance set at <0.05.

| RESULTS
A summary of the studies included in this systematic review including priming regimen, GHST used, cohort size, population characteristics and results can be seen in Table 1. There was significant heterogeneity in the included study populations, with children in the pre-and peripubertal age range (range 1-17 years), Tanner stage 1-5, variation in GHST used (ITT, arginine, glucagon, clonidine, levodopa, ornithine, exercise) and priming regimen used (oral oestrogen, transdermal oestrogen, intramuscular testosterone, mixed regimens). A detailed description of all of the studies can be found in Supporting Information: Table 5, and significant differences between the study populations described are highlighted in the results and discussion.
Unless otherwise stated a peak GH of 7 μg/L or higher was used as a normal response to GHST in all studies.

| Comparison of GH response with and without sex steroid priming: Randomised control trial (RCTs)
Four RCTs of GH response to stimulation with and without priming were identified. Results of these are summarised in Figure 2.
Martinez and colleagues randomised 15 prepubertal children with GHD and 44 short non-GHD children to oestrogen priming or placebo before sequential arginine-clonidine test. There was an increased GH response in primed non-GHD children versus placebo, but no significant effect of priming on GH response in the GHD group.
Priming improved the discrimination power of GHST between short non-GHD and GHD children. 12 Mean peak GH response following GHST was 17.8 ± 10.9 µg/L in the placebo group and 27.9 ± 14.5 µg/L in the oestrogen primed group in non-GHD children (p = .0001).  Figure 3. These can be divided into cohort studies of patient that underwent either primed or unprimed GHST, and studies where subjects underwent an unprimed GHST followed by a primed GHST.

T A B L E 1 Summary of studies included in the review
Three studies compared GH response between groups of patients that had undergone a single GHST, either primed or unprimed. Moll  | 531 that priming increased mean peak GH response (22 vs. 14.7 µg/L; p < .05) upon clonidine GHST. 27 Gonc and colleagues assessed 84 prepubertal and early-pubertal short boys who had previously not reached a GH of 10 µg/L on GHST. Priming increased mean peak GH response to levodopa following low dose testosterone (11.5 ± 8.8 µg/L vs. 4.8 ± 2.8 µg/L; p < .001) and conventional dose testosterone (13.0 ± 8.3 µg/L vs. 4.8 ± 2.5 µg/L; p < .001). 32 Muller and colleagues repeated arginine GH stimulation testing in 26 boys with short stature and delayed puberty, who had previously had low unprimed GHST responses (defined as peak GH <10 µg/L). Priming with testosterone increased mean peak GH response to arginine stimulation to a level above cut-off (10 µg/L) in 77% of patients who had previously had an insufficient response (16.8 ± 5.8 µg/L primed vs. 5.6 ± 2.6 µg/L unprimed; p < .05). 26 Molina and colleagues primed 39 children (with testosterone or oestradiol valerate) with delayed puberty and short stature, who had all previously had a low GH response to clonidine (peak GH <10 µg/L).
Priming increased mean peak GH response to clonidine stimulation (12.3 ± 8.7 µg/L vs. 4.9 ± 2.7 µg/L; p < .05). 18 Chatterjee et al. 23 assessed 28 boys with β-thalassaemia with pubertal failure and found that testosterone priming did not increase mean peak GH response to insulin tolerance test (

| Comparison of GH response with and without sex steroid priming: All studies
The composite results of RCTs, non-RCTs and all studies are presented in Figure 4, with data from all studies combined showing higher peak GH response in the primed group (p < .05).

| Effect of sex steroid dose used for priming on GH response to stimulation
Three studies included an evaluation of the number of doses used during priming on GH response to stimulation testing. Supporting Information: Table 6 provides a summary of the findings, the priming regimen, GHST used, cohort size and population characteristics of each study.
Gonc and colleagues assessed 21 prepubertal or early pubertal boys who had failed to exceed a peak GH above 10 µg/L upon single dose (low or conventional dose) testosterone priming. They found that multiple dose testosterone priming before repeat levodopa  Conflicting results, as to the effects of priming on GHST, have also been shown. 23,25,28 The two RCTs showing no significant difference between peak GH response both utilized clonidine for GH stimulation, raising the question as to whether effect of priming varies depending on GHST used. However, further studies 27 did find a significant increase in peak GH following primed clonidine tests.

| Auxological outcomes in primed versus unprimed individuals
Interestingly, the dose of testosterone utilised by Soliman et al. was the lowest seen in all studies included in this review (Supporting Information: Table 5), which may offer an explanation for their results showing no significant effect of testosterone priming. 28 The three studies addressing dosing regimen for sex steroid priming all concluded that the GH response to GHST was significantly increased with the higher dose sex steroid priming. 13,31,32 The age of children included in the Soliman et al study was slightly younger than those in other cohorts, and given that advancing Tanner stage alone was shown by Marin and colleagues to increase the percentage of children reaching adequate peak GH response, this is also likely to have influenced differential effect of priming in this study.
Chatterjee et al. 23 also showed that testosterone priming had no effect on GH response upon GHST, but was the only study to investigate boys with chronic disease (β-thalassaemia with pubertal failure), limiting the relevance of this data with respect to clinical practice.
Overall, studies that showed an effect of priming tended to be a higher level of evidence with low-moderate risk of bias. No studies reported a reduction in GH response following priming and no significant adverse effects of priming were reported in these studies.
Therefore, it can be concluded the priming before GH stimulation testing had the potential to increase peak GH result sufficiently to alter test outcome and subsequent clinical management.
It is important to consider the impact that priming before GHST has on auxological outcomes. For example, does priming allow patients who would benefit from GH treatment to be identified, and those that would not benefit (e.g., CDGP) to be appropriately not | 533 children who will respond to GH treatment in terms of increase in final height. Future studies to determine optimal priming regimen(s) would be beneficial. Consensus on a widely used and accepted peak GH cut-off response would also be beneficial. We would recommend subjects in future prospective studies should also be followed up to final height to allow assessment of impact on auxological outcomes.