- Email: [email protected]
Two years in growth hormone 2017–18
Growth Hormone & IGF Research 48–49 (2019) 60–64
Contents lists available at ScienceDirect
Growth Hormone & IGF Research journal homepage: www.elsevier.com/locate/ghir
Review Article
Two years in growth hormone 2017–18
T
⁎
P.E. Clayton , A.J. Whatmore Developmental Biology & Medicine, Faculty of Biology Medicine & Health, Manchester NIHR Academic Health Science Centre, University of Manchester, United Kingdom
A B S T R A C T
This brief review highlights new studies in three areas of the GH field, namely diagnostics, therapeutics and biomarkers. The diagnosis of GH deficiency has always presented a challenge: there is no “gold standard” test of GH status, and GH levels during stimulation testing are affected by many factors that limit diagnostic accuracy. Two new approaches to diagnosis have been proposed: one involves a classical endocrine test of GH production using a GH secretagogue to test the Ghrelin axis, and shows promise in the diagnosis of adult GH deficiency. The other uses a completely different approach analysing the individual's gene expression profile as a surrogate for GH status with high levels of test accuracy. From the therapeutic aspect, there have been significant efforts to produce a long-acting (LA) GH on the premise that this will improve adherence and patient convenience. Aspects of LA-GH pharmacology are considered, and it will be interesting to see in future years what place LA-GH GH takes in the market. Long term surveillance is a vital part of therapeutics; recent studies across Europe have provided reassurance on the safety of recombinant human GH (r-hGH) for those with uncomplicated growth disorders, but do emphasise the need to continue observation through adulthood. The search for biomarkers that precisely reflect GH action in children and adults is an ongoing task. One of the newer bone markers that shows promise is a fragment of collagen type X which now requires further investigation in humans. In parallel with the diagnostic studies, gene expression profiles at the start of r-hGH treatment have been used to predict GH response in children with GHD and girls with Turner syndrome. These data are promising but need evaluation across a range of growth disorders. R-hGH is an effective, safe therapy used in both children and adults. There is however a need to continue to refine diagnosis, treatment and most importantly longterm pharmacovigilance to ensure that the right patients have the best treatment with robust safety profiles.
1. Introduction Growth hormone plays a fundamental role in many body systems, and therefore is of interest to a wide range of scientists and clinicians. Using the simple term “Growth hormone” and the year in a PubMed search generates in the region of 1500 publications per year. The choice of material to illustrate what is happening in the field is therefore immense. This brief review, based on a lecture given at the 9th International Congress of the GH Research Society and the IGF Society in 2018, focuses on clinical advances in Diagnostics, Therapeutics and Biomarkers. In each of these areas important advances have been made in the last two years that are capturing new approaches to management. 1.1. Diagnostics The clinical characterization of an individual's GH status is complex and controversial [1]. The classical approach is to evaluate the capacity for release of GH from the pituitary in response to a stimulating agent (i.e. a pharmacological test) combined with a test of the downstream effector system for GH, namely IGF-I and its major serum binding protein IGFBP-3. There are many limitations to this approach - there is
no ‘gold standard’ test; stimulation testing ignores physiological GH release and will miss those with adequate GH reserve but inadequate daily secretion (so called neurosecretory dysfunction); GH release during stimulation tests is negatively influenced by body fat and positively by sex steroids; stimulation testing has a recognised high false positive rate (~20% of tests falsely indicate GH deficiency); and the IGF axis does not completely reflect GH status, but is influenced by nutritional status and other hormonal milieu. The introduction of a new approach to GH testing, that may help to reconcile some of the above issues, is therefore a welcome development. It has been recognised for many years that GH secretion can be stimulated by synthetic molecules, some of which were orally active, inferring the existence of an endogenous GH secretagogue receptor and a natural ligand. Ghrelin and its receptor are now well characterised [2]. This test therefore works through the GH secretagogue receptor, rather than the GH releasing hormone (GHRH) / somatostatin (SMS) axes. Garcia et al. [3] have examined the use of macimorelin (Mac), an oral GH secretagogue, as a diagnostic test for adult GH deficiency (AGHD). Using 4 groups of participants – high likelihood of A-GHD, intermediate likelihood, low likelihood and healthy controls – the GH
⁎ Corresponding author at: Royal Manchester Children's Hospital, 5th Floor (Research), Manchester University Hospital Foundation NHS Trust, Oxford Road, Manchester M13 9WL, United Kingdom E-mail address: [email protected] (P.E. Clayton).
https://doi.org/10.1016/j.ghir.2019.10.003 Received 3 October 2019; Received in revised form 14 October 2019; Accepted 14 October 2019 Available online 24 October 2019 1096-6374/ © 2019 Elsevier Ltd. All rights reserved.
Growth Hormone & IGF Research 48–49 (2019) 60–64
P.E. Clayton and A.J. Whatmore
Fig. 1. Macimorelin as a diagnostic test for Adult GHD: Peak GH Responses in the MAC and Insulin tolerance tests in 3 groups of patients with Adult GH deficiency (AGHD) and a group of Healthy controls - AGHD likelihood: Group A = High; B = Intermediate; C = Low; D = Healthy. Reprinted from Garcia et al, 2018 (ref 3) with permission.
Fig. 2. Sensitivity & Specificity of the Macimorelin GH Stimulation test over a range of peak GH cut-offs in those with a high likelihood of Adult GHD versus Healthy controls. Reprinted from Garcia et al, 2018 (ref 3) with permission.
with a bias to higher peaks in the Mac test. A performance evaluation using the high likelihood group versus healthy controls generated high sensitivity and specificity (Fig. 2). Stimulating the ghrelin axis adds a new dimension to testing, and its value when applied more widely
release in response to Mac has been measured and compared to that released in an insulin tolerance test (Fig. 1). In both tests those with high & intermediate likelihood of A-GHD had lower values than the low likelihood group and the controls. The two tests were well correlated 61
Growth Hormone & IGF Research 48–49 (2019) 60–64
P.E. Clayton and A.J. Whatmore
Peak GH (mcg/L)
2
4
6
-2
10
Normal
+2
Fig. 3. Gene expression in control and GHD children. This heat map shows individual patient gene expression profiles (as the columns) ranked by peak GH level in a GH stimulation test, while each row represents the level of expression of an individual mRNA. The genes were identified as those that correlated with peak GH response (both positively and negatively). The profiles of aged matched normal children are shown to the right. Random forest analysis for predicting GHD versus Normal children generated an Area under the Curve of 0.98 (95%CI 0.96–1.0) with a sensitivity of 100% and a specificity of 96%. On the heat map, Red indicates a gene with high expression. Green indicates a gene with low expression. Reprinted from Murray et al, 2018 (ref 4) with permission.
2. Therapeutics
including in the diagnosis of GHD in children will be interesting. In contrast to the longstanding hormonal approaches to the diagnosis of GHD, Murray et al. [4] have explored a completely new direction, namely using an individual's genomic signature as a marker of GH status. The test is based on the evaluation of mRNA expression profiles in blood cells, as a surrogate marker of the endogenous activity within the GH axis. White blood cells do express GH receptors [5] and show responses to GH stimulation. Murray et al. were able to show that the level of expression of 347 probesets on the mRNA array was correlated to peak GH concentration in a standard stimulation test (188 positively and 159 negatively) in a group of prepubertal GHD children, naïve to GH treatment, in the PREDICT study [4]. This infers that white blood cell gene expression does reflect the GH milieu. Interestingly, 65 of these probesets were represented in growth plate expression profiles. A small number of gene expression profiles from healthy children were obtained from online databases: using the two datasets in a random forest algorithm [Fig. 3], it was possible to predict GHD status with a sensitivity of 100% and a specificity of 96%, a performance that is significantly higher than hormonal testing. The transcriptome data could also be used to accurately predict severity of GHD – peak GH ≤4 μg/L versus 5 to < 10 μg/L. Pathway analysis on these GH-associated genes indicated a regulatory role for APC, SOX2, PIK3R3 and SIRT2, associated with WNT and Hedgehog signalling, myogenesis, mitochondrial biogenesis, and circadian clock. This test needs validation and testing in a full range of short stature scenarios, however it may open up a new diagnostic approach.
2.1. Long-acting GH Recombinant human GH (r-hGH) has been used worldwide since the mid-1980s. It is administered as a single subcutaneous injection on a daily basis; it has proved very effective in restoring normal growth in children with GHD with a good safety profile, and in improving the features associated with adult GHD. GH therapy has also been widely used in non-GHD short stature conditions, both safely and with improvement to predicted adult stature (in for instance the short child born small for gestational age and Turner girls) [6]. Daily injections are a burden and hence adherence can be a major issue diminishing efficacy. Over the last 20 years, pharma has made considerable efforts to develop a longer-acting GH preparation that can be used weekly or at longer intervals. Currently there are multiple preparations being tested across the world at various stages of development in healthy adults and in adult and childhood GHD. Modifications to the GH molecule to increase its longevity in human studies have included: fusion to antibody Fc fragment, pegylation, binding to albumen, formulation as a prodrug and C-terminal modification [7]. To date, one product has a license in some countries. Scientists continue to explore yet more avenues with a recent study looking at PEG-PLA nanoparticles as novel carriers for delivery of r-hGH [8]. Long-acting GHs have a very different pharmacokinetic profile compared to that generated by daily GH injections. There are therefore many questions that need to be answered in the development of these products. A selection includes: are rhythms in GH secretion important for effective action; what is the impact on growth versus metabolic 62
Growth Hormone & IGF Research 48–49 (2019) 60–64
P.E. Clayton and A.J. Whatmore
-2
+2
Fig. 4. Gene Expression profiles at baseline in prepubertal GHD children. This heat map is ranked by 1st year Growth Velocity (x-axis, cm/yr) in response to rhGH. Each column represents an individual patient, while each row represents the level of expression of an individual mRNA. The genes were identified as those that correlated with growth response (both positively and negatively). The yellow box includes those in the lowest quartile for growth response. There is a clear difference in their profile from those with higher growth responses. On the heat map, Red indicates a gene with high expression. Green indicates a gene with low expression. Reprinted from Clayton et al, 2013 (ref 20) with permission.
receiving radiotherapy, but the risk was not related to any GH-related parameter [14]. Metacohorts such as SAGhE have their limitations; for instance, there are no direct comparison untreated cohorts, there is insufficient control for selection bias and detailed data on r-hGH exposure and IGF-I levels are not available. However, in countries where such cohorts can be followed long-term, it is certainly worthwhile doing so. This will be equally relevant in future years to the surveillance of those on longacting r-hGH.
action; will adherence improve; will safety issues differ from daily rhGH [9,10]. A number of products have been discontinued due to lack of equivalence to daily r-hGH, side-effects or manufacturing issues, but others are continuing through the development pipeline. It will be a number of years yet before we know how frequently these products will be used in routine clinical practice.
2.2. Safety of r-hGH During the therapy years, r-hGH is considered to have a good safety profile. There is however evidence from in-vitro cell models, in-vivo animal studies and some human epidemiological surveys that links the GH-IGF axis to cancer [11]. Reassuringly, there is no evidence that during clinical use r-hGH causes the development of cancer [11]. Nevertheless, long-term safety surveillance in the years after GH therapy has been advocated, and has been addressed by studies such as SAGhE (Safety and Appropriateness of Growth hormone treatments in Europe). A meta-cohort of ~25,000 individuals treated with r-hGH during childhood and now into adulthood in 8 European countries has been assessed for incidence and mortality associated with major system disorders, including cancer, against background age-related population rates [12]. The median age of this cohort at the time of evaluation was ~30 years with the oldest in their late 30s. With respect to cancer, the analysis did not ‘generally support’ a role for r-hGH in cancer. However, there were some findings that indicate that ongoing surveillance should occur. These included an upward trend in cancer mortality in relation to r-hGH dose in those with a previous cancer and possible effects on bone cancer, bladder cancer and Hodgkin's lymphoma [13]. The latter observations were not seen in those with a diagnosis of isolated growth failure, but in those with a more complex, non-cancer associated growth disorder. Meningiomas were relatively common in those
2.3. Biomarkers When initiating a course of treatment, it is helpful to be able to counsel the patient on what response might be expected. For r-hGH treatment, which may be required for many years, the response in the first year determines the outcome in all subsequent years. So a baseline biomarker as well as an early response biomarker would certainly help in managing expectation. In growth disorders, models based on baseline characteristics have been developed that predict first year and later year's growth responses [15–17]. These are based on such parameters as degree of short stature and/or discrepancy from parental target, birth size, current weight, GH dose and in GHD peak GH during stimulation testing. These models are not commonly used in clinical practice. Individual biomarkers such as serum IGF-I, checked during treatment, have some value, but changes are influenced by many other factors than just r-hGH treatment, and therefore the correlation to growth response is relatively weak. Proteomic approaches to identifying markers for both GH deficiency and excess have met with some success, but have not translated into general clinical use [18]. Markers of bone turnover have been looked at over the years, and a recent study [19] has focussed on a degradation fragment of type X 63
Growth Hormone & IGF Research 48–49 (2019) 60–64
P.E. Clayton and A.J. Whatmore
collagen. Serum levels mirror growth in mice and the normal child growth velocity trajectory. The performance of this analyte should be explored in growth disorders. In parallel with the genomic approach to identifying GHD, gene expression profiles can be used to predict response to GH therapy, with two conditions evaluated to date – childhood GHD and Turner syndrome (TS) [20]. At baseline, genes, whose expression is related to first year growth response on r-hGH, can be used to identify those in the lowest quartile of response (Fig. 4). In addition, it is possible to use a gene network approach to identify a subset of genes expressed in both GHD and TS that predicts both first year and long-term response to rhGH treatment [21]. Such novel approaches need extensive validation, but they do offer an alternative, potentially more accurate method to predict outcomes, and introduce a ‘personalised’ approach to therapy. Mouse models have been used extensively to define the consequences of disrupting the GH-IGF axis either globally or in specific systems. Such work has provided and will continue to provide extremely important insights into mechanism and consequences. These models are also used to search for biomarkers. In recent years, the zebrafish has been looked at as a model to study the effects of chemicals on the GH-IGF axis [22] and the generation of recombinant ZF GH opens the opportunity for treatment studies [23].
93247. [5] J.-L. Bresson, S.B. Jeay, M.-C. Gagnerault, C. Kayser, N. Beressi, Z. Wu, ... M.C. Postel-Vinay, Growth hormone (gh) and prolactin receptors in human peripheral blood mononuclear cells: relation with age and gh-binding protein, Endocrinology 140 (7) (1999) 3203–3209, https://doi.org/10.1210/endo.140.7.6854%J Endocrinology. [6] R. Pfaffle, Hormone replacement therapy in children: the use of growth hormone and IGF-I, Best Pract. Res. Clin. Endocrinol. Metab. 29 (3) (2015) 339–352, https:// doi.org/10.1016/j.beem.2015.04.009. [7] K.C.J. Yuen, B.S. Miller, B.M.K. Biller, The current state of long-acting growth hormone preparations for growth hormone therapy, Curr. Opin. Endocrinol. Diabetes Obes. 25 (4) (2018) 267–273, https://doi.org/10.1097/med. 0000000000000416. [8] R. Ghasemi, M. Abdollahi, E. Emamgholi Zadeh, K. Khodabakhshi, A. Badeli, H. Bagheri, S. Hosseinkhani, mPEG-PLA and PLA-PEG-PLA nanoparticles as new carriers for delivery of recombinant human Growth Hormone (rhGH), Sci. Rep. 8 (1) (2018) 9854, https://doi.org/10.1038/s41598-018-28092-8. [9] J.S. Christiansen, P.F. Backeljauw, M. Bidlingmaier, B.M. Biller, M.C. Boguszewski, F.F. Casanueva, ... K. Yuen, Growth hormone research society perspective on the development of long-acting growth hormone preparations, Eur. J. Endocrinol. 174 (6) (2016) C1–C8, https://doi.org/10.1530/eje-16-0111. [10] C. Hoybye, P. Cohen, A.R. Hoffman, R. Ross, B.M. Biller, J.S. Christiansen, S. Growth Hormone Research, Status of long-acting-growth hormone preparations– 2015, Growth Hormon. IGF Res. 25 (5) (2015) 201–206, https://doi.org/10.1016/j. ghir.2015.07.004. [11] P.E. Clayton, I. Banerjee, P.G. Murray, A.G. Renehan, Growth hormone, the insulinlike growth factor axis, insulin and cancer risk, Nat. Rev. Endocrinol. 7 (1) (2011) 11–24, https://doi.org/10.1038/nrendo.2010.171. [12] A.J. Swerdlow, R. Cooke, K. Albertsson-Wikland, B. Borgstrom, G. Butler, S. Cianfarani, ... J.C. Carel, Description of the SAGhE cohort: a large European study of mortality and cancer incidence risks after childhood treatment with recombinant growth Hormone, Horm. Res. Paediatr. 84 (3) (2015) 172–183, https://doi.org/10. 1159/000435856. [13] A.J. Swerdlow, R. Cooke, D. Beckers, B. Borgstrom, G. Butler, J.C. Carel, ... G.R.J. Zandwijken, Cancer risks in patients treated with Growth Hormone in childhood: the SAGhE European cohort study, J. Clin. Endocrinol. Metab. 102 (5) (2017) 1661–1672, https://doi.org/10.1210/jc.2016-2046. [14] A.J. Swerdlow, R. Cooke, D. Beckers, G. Butler, J.C. Carel, S. Cianfarani, et al., Risk of meningioma in European patients treated with Growth Hormone in childhood: results from the SAGhE cohort, J. Clin. Endocrinol. Metab. 104 (2019) 658–664. [15] M.B. Ranke, A. Lindberg, P. Chatelain, P. Wilton, W. Cutfield, K. AlbertssonWikland, D.A. Price, Derivation and validation of a mathematical model for predicting the response to exogenous recombinant human growth hormone (GH) in prepubertal children with idiopathic GH deficiency. KIGS international board. Kabi pharmacia international growth study, J. Clin. Endocrinol. Metab. 84 (4) (1999) 1174–1183, https://doi.org/10.1210/jcem.84.4.5634. [16] M.B. Ranke, A. Lindberg, P. Chatelain, P. Wilton, W. Cutfield, K. AlbertssonWikland, D.A. Price, Predicting the response to recombinant human growth hormone in turner syndrome: KIGS models. KIGS international board. Kabi international growth study, Acta Paediatr. Suppl. 88 (433) (1999) 122–125. [17] M.B. Ranke, A. Lindberg, K.I. Board, Observed and predicted growth responses in prepubertal children with growth disorders: guidance of growth hormone treatment by empirical variables, J. Clin. Endocrinol. Metab. 95 (3) (2010) 1229–1237, https://doi.org/10.1210/jc.2009-1471. [18] S. Duran-Ortiz, A.L. Brittain, J.J. Kopchick, The impact of growth hormone on proteomic profiles: a review of mouse and adult human studies, Clin. Proteomics 14 (2017) 24, https://doi.org/10.1186/s12014-017-9160-2. [19] R.F. Coghlan, J.A. Oberdorf, S. Sienko, M.D. Aiona, B.A. Boston, K.J. Connelly, ... W.A. Horton, A degradation fragment of type X collagen is a real-time marker for bone growth velocity, Sci. Transl. Med. 9 (419) (2017), https://doi.org/10.1126/ scitranslmed.aan4669. [20] P. Clayton, P. Chatelain, L. Tato, H.W. Yoo, G.R. Ambler, A. Belgorosky, ... C. Olivier, A pharmacogenomic approach to the treatment of children with GH deficiency or turner syndrome, Eur. J. Endocrinol. 169 (3) (2013) 277–289, https:// doi.org/10.1530/EJE-13-0069. [21] A. Stevens, P. Murray, C. De Leonibus, T. Garner, E. Koledova, G. Ambler, ... P. Clayton, Gene expression signatures predict response to therapy with growth hormone, J bioRxiv (2019), https://doi.org/10.1101/637892. [22] Y. Dang, F. Wang, C. Liu, Real-time PCR array to study the effects of chemicals on the growth hormone/insulin-like growth factors (GH/IGFs) axis of zebrafish embryos/larvae, Chemosphere 207 (2018) 365–376, https://doi.org/10.1016/j. chemosphere.2018.05.102. [23] E. Oclon, G. Solomon, Z. Hayouka, A. Gertler, Preparation of biologically active monomeric recombinant zebrafish (Danio rerio) and rainbow trout (Oncorhynchus mykiss) recombinant growth hormones, Fish Physiol. Biochem. 44 (4) (2018) 1215–1222, https://doi.org/10.1007/s10695-018-0512-2.
3. Conclusion r-hGH has been in use for 35 years and has proved to be a safe therapy, that has enhanced the lives of many thousands of children with severe growth disorders and adults with GHD. The GH field continues to attract pharma investment with a significant interest in developing new products, in particular long-acting GHs. Over the last 2 years, clinical research has focussed on new approaches to diagnosis in GHD and new ways to identify markers of GH action as well as good from poor responders to r-hGH. Most importantly there is active interest in ensuring that long-term pharmacovigilance of all those that have received r-hGH at any stage in their lives should occur. Declaration of Competing Interest Peter Clayton is a member of the. Scientific Advisory Board, Lumos Pharma, Austin, Texas US. Acknowledgements All Figures are presented with permission from the original journals: Figs. 1 and 2 are from Garcia et al (2018; ref. 3), Fig. 3 is from Murray et al (2018; ref. 4) and Fig. 4 is from Clayton et al (2013; ref. 20) References [1] P.G. Murray, M.T. Dattani, P.E. Clayton, Controversies in the diagnosis and management of growth hormone deficiency in childhood and adolescence, Arch. Dis. Child 101 (1) (2016) 96–100, https://doi.org/10.1136/archdischild-2014-307228. [2] M. Kojima, K. Kangawa, Ghrelin: structure and function, Physiol. Rev. 85 (2) (2005) 495–522, https://doi.org/10.1152/physrev.00012.2004. [3] J.M. Garcia, B.M.K. Biller, M. Korbonits, V. Popovic, A. Luger, C.J. Strasburger, ... K.C.J. Yuen, Macimorelin as a diagnostic test for adult GH deficiency, J. Clin. Endocrinol. Metab. 103 (8) (2018) 3083–3093, https://doi.org/10.1210/jc.201800665. [4] P.G. Murray, A. Stevens, C. De Leonibus, E. Koledova, P. Chatelain, P.E. Clayton, Transcriptomics and machine learning predict diagnosis and severity of growth hormone deficiency, JCI Insight 3 (7) (2018), https://doi.org/10.1172/jci.insight.
64
Contents lists available at ScienceDirect
Growth Hormone & IGF Research journal homepage: www.elsevier.com/locate/ghir
Review Article
Two years in growth hormone 2017–18
T
⁎
P.E. Clayton , A.J. Whatmore Developmental Biology & Medicine, Faculty of Biology Medicine & Health, Manchester NIHR Academic Health Science Centre, University of Manchester, United Kingdom
A B S T R A C T
This brief review highlights new studies in three areas of the GH field, namely diagnostics, therapeutics and biomarkers. The diagnosis of GH deficiency has always presented a challenge: there is no “gold standard” test of GH status, and GH levels during stimulation testing are affected by many factors that limit diagnostic accuracy. Two new approaches to diagnosis have been proposed: one involves a classical endocrine test of GH production using a GH secretagogue to test the Ghrelin axis, and shows promise in the diagnosis of adult GH deficiency. The other uses a completely different approach analysing the individual's gene expression profile as a surrogate for GH status with high levels of test accuracy. From the therapeutic aspect, there have been significant efforts to produce a long-acting (LA) GH on the premise that this will improve adherence and patient convenience. Aspects of LA-GH pharmacology are considered, and it will be interesting to see in future years what place LA-GH GH takes in the market. Long term surveillance is a vital part of therapeutics; recent studies across Europe have provided reassurance on the safety of recombinant human GH (r-hGH) for those with uncomplicated growth disorders, but do emphasise the need to continue observation through adulthood. The search for biomarkers that precisely reflect GH action in children and adults is an ongoing task. One of the newer bone markers that shows promise is a fragment of collagen type X which now requires further investigation in humans. In parallel with the diagnostic studies, gene expression profiles at the start of r-hGH treatment have been used to predict GH response in children with GHD and girls with Turner syndrome. These data are promising but need evaluation across a range of growth disorders. R-hGH is an effective, safe therapy used in both children and adults. There is however a need to continue to refine diagnosis, treatment and most importantly longterm pharmacovigilance to ensure that the right patients have the best treatment with robust safety profiles.
1. Introduction Growth hormone plays a fundamental role in many body systems, and therefore is of interest to a wide range of scientists and clinicians. Using the simple term “Growth hormone” and the year in a PubMed search generates in the region of 1500 publications per year. The choice of material to illustrate what is happening in the field is therefore immense. This brief review, based on a lecture given at the 9th International Congress of the GH Research Society and the IGF Society in 2018, focuses on clinical advances in Diagnostics, Therapeutics and Biomarkers. In each of these areas important advances have been made in the last two years that are capturing new approaches to management. 1.1. Diagnostics The clinical characterization of an individual's GH status is complex and controversial [1]. The classical approach is to evaluate the capacity for release of GH from the pituitary in response to a stimulating agent (i.e. a pharmacological test) combined with a test of the downstream effector system for GH, namely IGF-I and its major serum binding protein IGFBP-3. There are many limitations to this approach - there is
no ‘gold standard’ test; stimulation testing ignores physiological GH release and will miss those with adequate GH reserve but inadequate daily secretion (so called neurosecretory dysfunction); GH release during stimulation tests is negatively influenced by body fat and positively by sex steroids; stimulation testing has a recognised high false positive rate (~20% of tests falsely indicate GH deficiency); and the IGF axis does not completely reflect GH status, but is influenced by nutritional status and other hormonal milieu. The introduction of a new approach to GH testing, that may help to reconcile some of the above issues, is therefore a welcome development. It has been recognised for many years that GH secretion can be stimulated by synthetic molecules, some of which were orally active, inferring the existence of an endogenous GH secretagogue receptor and a natural ligand. Ghrelin and its receptor are now well characterised [2]. This test therefore works through the GH secretagogue receptor, rather than the GH releasing hormone (GHRH) / somatostatin (SMS) axes. Garcia et al. [3] have examined the use of macimorelin (Mac), an oral GH secretagogue, as a diagnostic test for adult GH deficiency (AGHD). Using 4 groups of participants – high likelihood of A-GHD, intermediate likelihood, low likelihood and healthy controls – the GH
⁎ Corresponding author at: Royal Manchester Children's Hospital, 5th Floor (Research), Manchester University Hospital Foundation NHS Trust, Oxford Road, Manchester M13 9WL, United Kingdom E-mail address: [email protected] (P.E. Clayton).
https://doi.org/10.1016/j.ghir.2019.10.003 Received 3 October 2019; Received in revised form 14 October 2019; Accepted 14 October 2019 Available online 24 October 2019 1096-6374/ © 2019 Elsevier Ltd. All rights reserved.
Growth Hormone & IGF Research 48–49 (2019) 60–64
P.E. Clayton and A.J. Whatmore
Fig. 1. Macimorelin as a diagnostic test for Adult GHD: Peak GH Responses in the MAC and Insulin tolerance tests in 3 groups of patients with Adult GH deficiency (AGHD) and a group of Healthy controls - AGHD likelihood: Group A = High; B = Intermediate; C = Low; D = Healthy. Reprinted from Garcia et al, 2018 (ref 3) with permission.
Fig. 2. Sensitivity & Specificity of the Macimorelin GH Stimulation test over a range of peak GH cut-offs in those with a high likelihood of Adult GHD versus Healthy controls. Reprinted from Garcia et al, 2018 (ref 3) with permission.
with a bias to higher peaks in the Mac test. A performance evaluation using the high likelihood group versus healthy controls generated high sensitivity and specificity (Fig. 2). Stimulating the ghrelin axis adds a new dimension to testing, and its value when applied more widely
release in response to Mac has been measured and compared to that released in an insulin tolerance test (Fig. 1). In both tests those with high & intermediate likelihood of A-GHD had lower values than the low likelihood group and the controls. The two tests were well correlated 61
Growth Hormone & IGF Research 48–49 (2019) 60–64
P.E. Clayton and A.J. Whatmore
Peak GH (mcg/L)
2
4
6
-2
10
Normal
+2
Fig. 3. Gene expression in control and GHD children. This heat map shows individual patient gene expression profiles (as the columns) ranked by peak GH level in a GH stimulation test, while each row represents the level of expression of an individual mRNA. The genes were identified as those that correlated with peak GH response (both positively and negatively). The profiles of aged matched normal children are shown to the right. Random forest analysis for predicting GHD versus Normal children generated an Area under the Curve of 0.98 (95%CI 0.96–1.0) with a sensitivity of 100% and a specificity of 96%. On the heat map, Red indicates a gene with high expression. Green indicates a gene with low expression. Reprinted from Murray et al, 2018 (ref 4) with permission.
2. Therapeutics
including in the diagnosis of GHD in children will be interesting. In contrast to the longstanding hormonal approaches to the diagnosis of GHD, Murray et al. [4] have explored a completely new direction, namely using an individual's genomic signature as a marker of GH status. The test is based on the evaluation of mRNA expression profiles in blood cells, as a surrogate marker of the endogenous activity within the GH axis. White blood cells do express GH receptors [5] and show responses to GH stimulation. Murray et al. were able to show that the level of expression of 347 probesets on the mRNA array was correlated to peak GH concentration in a standard stimulation test (188 positively and 159 negatively) in a group of prepubertal GHD children, naïve to GH treatment, in the PREDICT study [4]. This infers that white blood cell gene expression does reflect the GH milieu. Interestingly, 65 of these probesets were represented in growth plate expression profiles. A small number of gene expression profiles from healthy children were obtained from online databases: using the two datasets in a random forest algorithm [Fig. 3], it was possible to predict GHD status with a sensitivity of 100% and a specificity of 96%, a performance that is significantly higher than hormonal testing. The transcriptome data could also be used to accurately predict severity of GHD – peak GH ≤4 μg/L versus 5 to < 10 μg/L. Pathway analysis on these GH-associated genes indicated a regulatory role for APC, SOX2, PIK3R3 and SIRT2, associated with WNT and Hedgehog signalling, myogenesis, mitochondrial biogenesis, and circadian clock. This test needs validation and testing in a full range of short stature scenarios, however it may open up a new diagnostic approach.
2.1. Long-acting GH Recombinant human GH (r-hGH) has been used worldwide since the mid-1980s. It is administered as a single subcutaneous injection on a daily basis; it has proved very effective in restoring normal growth in children with GHD with a good safety profile, and in improving the features associated with adult GHD. GH therapy has also been widely used in non-GHD short stature conditions, both safely and with improvement to predicted adult stature (in for instance the short child born small for gestational age and Turner girls) [6]. Daily injections are a burden and hence adherence can be a major issue diminishing efficacy. Over the last 20 years, pharma has made considerable efforts to develop a longer-acting GH preparation that can be used weekly or at longer intervals. Currently there are multiple preparations being tested across the world at various stages of development in healthy adults and in adult and childhood GHD. Modifications to the GH molecule to increase its longevity in human studies have included: fusion to antibody Fc fragment, pegylation, binding to albumen, formulation as a prodrug and C-terminal modification [7]. To date, one product has a license in some countries. Scientists continue to explore yet more avenues with a recent study looking at PEG-PLA nanoparticles as novel carriers for delivery of r-hGH [8]. Long-acting GHs have a very different pharmacokinetic profile compared to that generated by daily GH injections. There are therefore many questions that need to be answered in the development of these products. A selection includes: are rhythms in GH secretion important for effective action; what is the impact on growth versus metabolic 62
Growth Hormone & IGF Research 48–49 (2019) 60–64
P.E. Clayton and A.J. Whatmore
-2
+2
Fig. 4. Gene Expression profiles at baseline in prepubertal GHD children. This heat map is ranked by 1st year Growth Velocity (x-axis, cm/yr) in response to rhGH. Each column represents an individual patient, while each row represents the level of expression of an individual mRNA. The genes were identified as those that correlated with growth response (both positively and negatively). The yellow box includes those in the lowest quartile for growth response. There is a clear difference in their profile from those with higher growth responses. On the heat map, Red indicates a gene with high expression. Green indicates a gene with low expression. Reprinted from Clayton et al, 2013 (ref 20) with permission.
receiving radiotherapy, but the risk was not related to any GH-related parameter [14]. Metacohorts such as SAGhE have their limitations; for instance, there are no direct comparison untreated cohorts, there is insufficient control for selection bias and detailed data on r-hGH exposure and IGF-I levels are not available. However, in countries where such cohorts can be followed long-term, it is certainly worthwhile doing so. This will be equally relevant in future years to the surveillance of those on longacting r-hGH.
action; will adherence improve; will safety issues differ from daily rhGH [9,10]. A number of products have been discontinued due to lack of equivalence to daily r-hGH, side-effects or manufacturing issues, but others are continuing through the development pipeline. It will be a number of years yet before we know how frequently these products will be used in routine clinical practice.
2.2. Safety of r-hGH During the therapy years, r-hGH is considered to have a good safety profile. There is however evidence from in-vitro cell models, in-vivo animal studies and some human epidemiological surveys that links the GH-IGF axis to cancer [11]. Reassuringly, there is no evidence that during clinical use r-hGH causes the development of cancer [11]. Nevertheless, long-term safety surveillance in the years after GH therapy has been advocated, and has been addressed by studies such as SAGhE (Safety and Appropriateness of Growth hormone treatments in Europe). A meta-cohort of ~25,000 individuals treated with r-hGH during childhood and now into adulthood in 8 European countries has been assessed for incidence and mortality associated with major system disorders, including cancer, against background age-related population rates [12]. The median age of this cohort at the time of evaluation was ~30 years with the oldest in their late 30s. With respect to cancer, the analysis did not ‘generally support’ a role for r-hGH in cancer. However, there were some findings that indicate that ongoing surveillance should occur. These included an upward trend in cancer mortality in relation to r-hGH dose in those with a previous cancer and possible effects on bone cancer, bladder cancer and Hodgkin's lymphoma [13]. The latter observations were not seen in those with a diagnosis of isolated growth failure, but in those with a more complex, non-cancer associated growth disorder. Meningiomas were relatively common in those
2.3. Biomarkers When initiating a course of treatment, it is helpful to be able to counsel the patient on what response might be expected. For r-hGH treatment, which may be required for many years, the response in the first year determines the outcome in all subsequent years. So a baseline biomarker as well as an early response biomarker would certainly help in managing expectation. In growth disorders, models based on baseline characteristics have been developed that predict first year and later year's growth responses [15–17]. These are based on such parameters as degree of short stature and/or discrepancy from parental target, birth size, current weight, GH dose and in GHD peak GH during stimulation testing. These models are not commonly used in clinical practice. Individual biomarkers such as serum IGF-I, checked during treatment, have some value, but changes are influenced by many other factors than just r-hGH treatment, and therefore the correlation to growth response is relatively weak. Proteomic approaches to identifying markers for both GH deficiency and excess have met with some success, but have not translated into general clinical use [18]. Markers of bone turnover have been looked at over the years, and a recent study [19] has focussed on a degradation fragment of type X 63
Growth Hormone & IGF Research 48–49 (2019) 60–64
P.E. Clayton and A.J. Whatmore
collagen. Serum levels mirror growth in mice and the normal child growth velocity trajectory. The performance of this analyte should be explored in growth disorders. In parallel with the genomic approach to identifying GHD, gene expression profiles can be used to predict response to GH therapy, with two conditions evaluated to date – childhood GHD and Turner syndrome (TS) [20]. At baseline, genes, whose expression is related to first year growth response on r-hGH, can be used to identify those in the lowest quartile of response (Fig. 4). In addition, it is possible to use a gene network approach to identify a subset of genes expressed in both GHD and TS that predicts both first year and long-term response to rhGH treatment [21]. Such novel approaches need extensive validation, but they do offer an alternative, potentially more accurate method to predict outcomes, and introduce a ‘personalised’ approach to therapy. Mouse models have been used extensively to define the consequences of disrupting the GH-IGF axis either globally or in specific systems. Such work has provided and will continue to provide extremely important insights into mechanism and consequences. These models are also used to search for biomarkers. In recent years, the zebrafish has been looked at as a model to study the effects of chemicals on the GH-IGF axis [22] and the generation of recombinant ZF GH opens the opportunity for treatment studies [23].
93247. [5] J.-L. Bresson, S.B. Jeay, M.-C. Gagnerault, C. Kayser, N. Beressi, Z. Wu, ... M.C. Postel-Vinay, Growth hormone (gh) and prolactin receptors in human peripheral blood mononuclear cells: relation with age and gh-binding protein, Endocrinology 140 (7) (1999) 3203–3209, https://doi.org/10.1210/endo.140.7.6854%J Endocrinology. [6] R. Pfaffle, Hormone replacement therapy in children: the use of growth hormone and IGF-I, Best Pract. Res. Clin. Endocrinol. Metab. 29 (3) (2015) 339–352, https:// doi.org/10.1016/j.beem.2015.04.009. [7] K.C.J. Yuen, B.S. Miller, B.M.K. Biller, The current state of long-acting growth hormone preparations for growth hormone therapy, Curr. Opin. Endocrinol. Diabetes Obes. 25 (4) (2018) 267–273, https://doi.org/10.1097/med. 0000000000000416. [8] R. Ghasemi, M. Abdollahi, E. Emamgholi Zadeh, K. Khodabakhshi, A. Badeli, H. Bagheri, S. Hosseinkhani, mPEG-PLA and PLA-PEG-PLA nanoparticles as new carriers for delivery of recombinant human Growth Hormone (rhGH), Sci. Rep. 8 (1) (2018) 9854, https://doi.org/10.1038/s41598-018-28092-8. [9] J.S. Christiansen, P.F. Backeljauw, M. Bidlingmaier, B.M. Biller, M.C. Boguszewski, F.F. Casanueva, ... K. Yuen, Growth hormone research society perspective on the development of long-acting growth hormone preparations, Eur. J. Endocrinol. 174 (6) (2016) C1–C8, https://doi.org/10.1530/eje-16-0111. [10] C. Hoybye, P. Cohen, A.R. Hoffman, R. Ross, B.M. Biller, J.S. Christiansen, S. Growth Hormone Research, Status of long-acting-growth hormone preparations– 2015, Growth Hormon. IGF Res. 25 (5) (2015) 201–206, https://doi.org/10.1016/j. ghir.2015.07.004. [11] P.E. Clayton, I. Banerjee, P.G. Murray, A.G. Renehan, Growth hormone, the insulinlike growth factor axis, insulin and cancer risk, Nat. Rev. Endocrinol. 7 (1) (2011) 11–24, https://doi.org/10.1038/nrendo.2010.171. [12] A.J. Swerdlow, R. Cooke, K. Albertsson-Wikland, B. Borgstrom, G. Butler, S. Cianfarani, ... J.C. Carel, Description of the SAGhE cohort: a large European study of mortality and cancer incidence risks after childhood treatment with recombinant growth Hormone, Horm. Res. Paediatr. 84 (3) (2015) 172–183, https://doi.org/10. 1159/000435856. [13] A.J. Swerdlow, R. Cooke, D. Beckers, B. Borgstrom, G. Butler, J.C. Carel, ... G.R.J. Zandwijken, Cancer risks in patients treated with Growth Hormone in childhood: the SAGhE European cohort study, J. Clin. Endocrinol. Metab. 102 (5) (2017) 1661–1672, https://doi.org/10.1210/jc.2016-2046. [14] A.J. Swerdlow, R. Cooke, D. Beckers, G. Butler, J.C. Carel, S. Cianfarani, et al., Risk of meningioma in European patients treated with Growth Hormone in childhood: results from the SAGhE cohort, J. Clin. Endocrinol. Metab. 104 (2019) 658–664. [15] M.B. Ranke, A. Lindberg, P. Chatelain, P. Wilton, W. Cutfield, K. AlbertssonWikland, D.A. Price, Derivation and validation of a mathematical model for predicting the response to exogenous recombinant human growth hormone (GH) in prepubertal children with idiopathic GH deficiency. KIGS international board. Kabi pharmacia international growth study, J. Clin. Endocrinol. Metab. 84 (4) (1999) 1174–1183, https://doi.org/10.1210/jcem.84.4.5634. [16] M.B. Ranke, A. Lindberg, P. Chatelain, P. Wilton, W. Cutfield, K. AlbertssonWikland, D.A. Price, Predicting the response to recombinant human growth hormone in turner syndrome: KIGS models. KIGS international board. Kabi international growth study, Acta Paediatr. Suppl. 88 (433) (1999) 122–125. [17] M.B. Ranke, A. Lindberg, K.I. Board, Observed and predicted growth responses in prepubertal children with growth disorders: guidance of growth hormone treatment by empirical variables, J. Clin. Endocrinol. Metab. 95 (3) (2010) 1229–1237, https://doi.org/10.1210/jc.2009-1471. [18] S. Duran-Ortiz, A.L. Brittain, J.J. Kopchick, The impact of growth hormone on proteomic profiles: a review of mouse and adult human studies, Clin. Proteomics 14 (2017) 24, https://doi.org/10.1186/s12014-017-9160-2. [19] R.F. Coghlan, J.A. Oberdorf, S. Sienko, M.D. Aiona, B.A. Boston, K.J. Connelly, ... W.A. Horton, A degradation fragment of type X collagen is a real-time marker for bone growth velocity, Sci. Transl. Med. 9 (419) (2017), https://doi.org/10.1126/ scitranslmed.aan4669. [20] P. Clayton, P. Chatelain, L. Tato, H.W. Yoo, G.R. Ambler, A. Belgorosky, ... C. Olivier, A pharmacogenomic approach to the treatment of children with GH deficiency or turner syndrome, Eur. J. Endocrinol. 169 (3) (2013) 277–289, https:// doi.org/10.1530/EJE-13-0069. [21] A. Stevens, P. Murray, C. De Leonibus, T. Garner, E. Koledova, G. Ambler, ... P. Clayton, Gene expression signatures predict response to therapy with growth hormone, J bioRxiv (2019), https://doi.org/10.1101/637892. [22] Y. Dang, F. Wang, C. Liu, Real-time PCR array to study the effects of chemicals on the growth hormone/insulin-like growth factors (GH/IGFs) axis of zebrafish embryos/larvae, Chemosphere 207 (2018) 365–376, https://doi.org/10.1016/j. chemosphere.2018.05.102. [23] E. Oclon, G. Solomon, Z. Hayouka, A. Gertler, Preparation of biologically active monomeric recombinant zebrafish (Danio rerio) and rainbow trout (Oncorhynchus mykiss) recombinant growth hormones, Fish Physiol. Biochem. 44 (4) (2018) 1215–1222, https://doi.org/10.1007/s10695-018-0512-2.
3. Conclusion r-hGH has been in use for 35 years and has proved to be a safe therapy, that has enhanced the lives of many thousands of children with severe growth disorders and adults with GHD. The GH field continues to attract pharma investment with a significant interest in developing new products, in particular long-acting GHs. Over the last 2 years, clinical research has focussed on new approaches to diagnosis in GHD and new ways to identify markers of GH action as well as good from poor responders to r-hGH. Most importantly there is active interest in ensuring that long-term pharmacovigilance of all those that have received r-hGH at any stage in their lives should occur. Declaration of Competing Interest Peter Clayton is a member of the. Scientific Advisory Board, Lumos Pharma, Austin, Texas US. Acknowledgements All Figures are presented with permission from the original journals: Figs. 1 and 2 are from Garcia et al (2018; ref. 3), Fig. 3 is from Murray et al (2018; ref. 4) and Fig. 4 is from Clayton et al (2013; ref. 20) References [1] P.G. Murray, M.T. Dattani, P.E. Clayton, Controversies in the diagnosis and management of growth hormone deficiency in childhood and adolescence, Arch. Dis. Child 101 (1) (2016) 96–100, https://doi.org/10.1136/archdischild-2014-307228. [2] M. Kojima, K. Kangawa, Ghrelin: structure and function, Physiol. Rev. 85 (2) (2005) 495–522, https://doi.org/10.1152/physrev.00012.2004. [3] J.M. Garcia, B.M.K. Biller, M. Korbonits, V. Popovic, A. Luger, C.J. Strasburger, ... K.C.J. Yuen, Macimorelin as a diagnostic test for adult GH deficiency, J. Clin. Endocrinol. Metab. 103 (8) (2018) 3083–3093, https://doi.org/10.1210/jc.201800665. [4] P.G. Murray, A. Stevens, C. De Leonibus, E. Koledova, P. Chatelain, P.E. Clayton, Transcriptomics and machine learning predict diagnosis and severity of growth hormone deficiency, JCI Insight 3 (7) (2018), https://doi.org/10.1172/jci.insight.
64