Growth and Growth Hormone: Current Applications and Clinical Updates

Introduction Advances in the Diagnosis of Growth and Growth Hormone-Related Deficiencies Ron G. Rosenfeld, MD Metabolic Consequences of Growth-Hormone Deficiency Mark A. Sperling, MD Dosing of Growth Hormone for theTreatment of Short Stature Pinchas Cohen, MD Emerging Applications of Growth-Hormone Therapies David B. Allen, MD Transition of Patients with Growth-Hormone Deficiency to Adult Care Nelly Mauras, MD

Introduction

Growth hormone (GH) deficiency affects nearly 20,000 children in the United States. GH deficiency may occur singularly or in combination with one or more pituitary hormone deficiencies. Short stature due to GH deficiency is a concern for patients and the families of patients with short stature, and the management of short stature can be difficult as a child transitions into adolescence and adulthood. New diagnostic techniques and applications of GH therapies may aid in successful management of GH deficiency. In addition, the relationship between GH and IGF-1 may allow clinicians to dose GH based on measurements of serum IGF-1 levels. The complex issues of GH deficiency present a clear need for education of healthcare providers for optimal treatment management.

To address some of these issues, an expert panel was convened at the Pediatric Academic Society (PAS) 2008 annual meeting in Honolulu, Hawaii. The panel shared their years of clinical experience to discuss the difficulties in accurate diagnosis, the metabolic consequences of GH deficiency, dosing strategies, emerging applications of GH, and the transition from childhood to adulthood. I thank the PAS, Vindico Medical Education, and the panel for their participation in the expert panel discussion and the development of this monograph.

—Mark A. Sperling, MD

Ron G. Rosenfeld, MD Senior Vice-President for Medical Affairs Lucile Packard Foundation for Children’s Health Professor of Pediatrics Stanford University Professor of Pediatrics Oregon Health and Science University Portland, OR. Mark A. Sperling, MD Professor of Pediatrics Department of Endocrinology, Diabetes and Metabolism Children’s Hospital University of Pittsburgh School of Medicine Pittsburgh, PA. Pinchas Cohen, MD Professor and Chief of Endocrinology Mattel Children’s Hospital at UCLA David Geffen School of Medicine Los Angeles, CA. David B. Allen, MD University of Wisconsin School of Medicine and Public Health Director of Endocrinology and Endocrinology & Diabetes Fellowship Training University of Wisconsin American Family Children’s Hospital Madison, WI. Nelly Mauras, MD Chief, Division of Endocrinology Pediatric Endocrinologist Department of Pediatrics Division of Endocrinology Nemours Children’s Clinic Jacksonville, FL.

Advances in the Diagnosis of Growth and Growth Hormone–Related Deficiencies
Ron G. Rosenfeld, MD

The problem in defining short stature is that any definition of short stature based on statistical criteria is arbitrary. Further, it is impossible to discriminate completely between the lower limits of normal height and the upper limits of pathology. The more that is learned about the genetic and genomic regulation of growth, the clearer it becomes that an extreme of normalcy cannot always be differentiated from pathologic growth. This fact makes the decision to treat short stature with any therapeutic agent quite difficult.

No matter how many children are treated, 5% will always be below the fifth percentile for height, and 1.2% will always lie below -2.25 standard deviations (SDs). Given this reality, it is neither practical nor desirable to treat every child with short stature. It must also be accepted that responses to treatment are highly variable. Treatment should be directed at children with the most severe growth failure and the greatest likelihood of responding to specific therapy.

The IGF-1 system is the major regulator of mammalian growth both in utero and postnatally. Whereas 20 or 30 years ago, the first question clinicians may have asked when confronted with a child of short stature was whether that child was growth hormone (GH)-deficient, it is now probably more important to be asking whether that child is IGF-deficient. One of the primary responsibilities is to determine if an IGF deficiency is due to a defect in GH production (secondary IGF deficiency) or if the defect is at the level of the GH receptor, the post-receptor signaling cascade, or the IGF-1 gene, itself (primary IGF deficiency).

In the case of secondary IGF deficiency, in which the abnormality is due to a defect in GH production, the treatment of choice is clearly GH.

If, however, the defect is at the level of the GH receptor, a signaling defect, or at the level of the IGF-1 gene, then several therapeutic options are available. For primary IGF deficiency, one of the difficulties is trying to decide whether a child should be treated with GH, IGF-1, or perhaps the combination of the two.

GH Stimulation Testing

To choose the best therapeutic option, one of the questions that must be answered is whether to perform a GH stimulation test. GH stimulation tests in this instance have been highly controversial. In my opinion, a GH stimulation test is strongly preferred for several reasons:

• It is the best method to differentiate GH deficiency from GH insensitivity.
• If the patient is GH deficient, it allows for evaluation of other pituitary functions, such as ACTH, gonadotrophin, or thyroid-stimulating hormone.
• It determines the need for magnetic resonance imaging, which is generally unnecessary if a child is not GH-deficient.
• It may be relevant genetically, as the decision to evaluate the child for a hereditary disorder rests on the underlying pathology.
• It predicts to a large extent, responsiveness to GH treatment.
• It guides management after the child completes his growth.

Although the advantages of performing a GH stimulation test are numerous, GH stimulation tests nevertheless need to be performed properly.

Marin and colleagues performed GH stimulation tests in 84 children with normal stature.¹ All subjects underwent a treadmill exercise test and an arginine-insulin provocative test, with and without sex steroid priming. In pubertal children, the peak GH increased with pubertal stage. One hundred percent of children who were Tanner 4 or 5, and 89% of those who were Tanner 3, achieved a GH level of at least 7 ng/ml.

In children who were Tanner 1, however, only 39% achieved a GH level of at least 7 ng/ml in the absence of sex steroid priming, which means that 61% of entirely normal children would have been identified as GH-deficient in the absence of sex steroid priming.

In pre-pubertal children without ethanol estradiol priming, the peak GH was as low as 1.9 ng/ml following exercise testing and arginine-insulin testing. With sex steroid priming in the same group of children, the peak GH ranged from 7.2 to 40.5 ng/ml.

These data argue that there is little value to performing any GH stimulation test in pre-pubertal children unless they are primed with sex steroids. In the absence of priming, the specificity of GH stimulation tests is highly suspect.

They also suggest that many children identified as idiopathic GH-deficient, who subsequently passed their tests when retested as adults, received an erroneous diagnosis of GH deficiency as children because the tests were not performed with sex steroid priming.

When to Perform MRI

The clinical presentation and laboratory findings should guide the decision to perform an MRI scan. There is little indication for performing an MRI on a short child with normal GH levels. Performing an MRI in lieu of a GH stimulation test subjects children who may not be GH deficient to an unnecessary, expensive, and often frightening test.

Gauging Response to Treatment

Criteria for defining satisfactory GH response have never been developed. The range of GH responsiveness is large; differences can be attributed to diagnosis, age, GH dose, parental height, compliance, inter-current illness, other therapies, and still poorly defined molecular and biochemical factors. Bakker and colleagues used data from National Cooperative Growth Study (NCGS) to construct plots of GH response during the first year of GH treatment with standard doses.² A hypothetical 5-year-old girl on GH therapy would have grown about 7 cm during her first year on therapy, which represents a -1.5 SD growth response (Figure 1). As proposed by the authors, a growth response below -1 SD should be considered a marginal response and warrant reevaluation of the child.

Figure 1: Females 1-14 Years with IGHD

Nonresponse to GH Therapy

Responsiveness should be checked regularly. If the child is not responding, investigate adherence, and recheck, if a satisfactory growth response continues to be absent. Look for inter-current illness. Consider the very rare possibility of antigrowth hormone antibodies. Re-evaluate the diagnosis of GH deficiency, especially if sex steroid priming was not used. Finally, consider the possibility that the child has primary IGF deficiency.

Some of the molecular etiologies of severe primary IGF deficiency are classic Laron syndrome, defects of the GH receptor in the transmembrane or intracellular domain, STAT-5b mutation (STAT-b is the critical link between the GH receptor and the IGF-1 gene), IGF-1 gene defects, and mutations in the gene for the acid labile subunit. All of these molecular etiologies add up to approximately 300 known causes, although these probably represent the tip of the iceberg for GH insensitivity (Figure 2).

Figure 2: GH - IGF-1 Axis

Options for Poorly Responsive Patients

What are the options for poorly responsive children? Treatment can be stopped, which is certainly an FDA-approved alternative.

A number of options that are not FDA-approved can be considered. The dosage of GH can be increased, based on auxology or IGF-1 concentrations. Therapy with IGF-1, or GH in combination with IGF-1, can also be considered. As another resort, epiphyseal fusion can be delayed. All these options should be considered on a case-by-case basis in a child who is not growing adequately in response to conventional dosages of GH.

References

1. Marin G, Domene HM, Barnes KM, Blackwell BJ, Cassoria FG, Cutler GB Jr. J Clin Endocrin Metab, 1994;79: 537-541.
2. Bakker B, Frane J, Anhalt H, Lippe B, Rosenfeld RG. J Clin Endocrin Metab, 2008;93:352-357.
3. Tauber M, Moulin P, Pienkowski C, Jouret B, Rochiccioli P. J Clin Endocrinol Metab, 1997;82:352-356.
4. Cutler L. Kenney GM. Arch Pediatric Adolec Med 2007; 161: 630-633

Metabolic Effects of Growth Hormone
Mark A. Sperling, MD

Growth hormone affects metabolism of protein, fat, and carbohydrate, which is the most profound during the pubertal increase in growth hormone secretion. Increased secretion of GHduring puberty produces a transient insulin resistance for carbohydrate metabolism and an accentuation of protein anabolism. In normal adolescents, increased insulin secretion compensates for the pubertal rise in insulin resistance. The increase in insulin synergizes with GH to enhance anabolism of protein. In addition, increases in GH enhance lipolysis and fat oxidation, sparing glucose and amino acids as fuel for growth. Sex steroids produced during puberty further enhance growth.

People who cannot compensate for pubertal-induced insulin resistance with an increase in insulin secretion can develop glucose intolerance or diabetes mellitus. Susceptible people may include those with a genetic predisposition to diabetes or with other factors that increase risk for diabetes such as obesity or an autoimmune assault on the pancreas. Understanding how GH affects metabolism gives insight into how adolescents acquire muscle, fat, and bone; the metabolic changes that occur during puberty; and the side effects that can manifest in susceptible people during GH therapy.

Effects of Growth Hormone on Metabolism

In addition to spurring growth, GH has direct effects on metabolism. In adipose tissue, it has catabolic effects, increasing lipolysis and decreasing glucose transport. It has anabolic effects on protein metabolism, producing increased amino acid transport, increased nitrogen retention, and resultant increases in lean muscle mass. Increases in GH during puberty affect carbohydrate metabolism by inducing insulin resistance. This induction of insulin resistance can unmask hyperglycemia in those who are genetically predisposed, as in people with the defects in KCNJ11 gene, TCFL7-like gene, or PPAR?. Growth hormone affects bone mineral metabolism by accelerating accretion of bone mineral, resulting in an increase in bone mass. (Figure 1).

Figure 1: Growth: Traditional View

Insulin Sensitivity and Response During Puberty

Oral glucose tolerance tests in adolescents at Tanner stage 3 or greater as in a pregnant women in their second or third trimester, shows that glucose tolerance is normal but insulin responses are increased. Bloch and colleagues demonstrated that decreased sensitivity to insulin occurs in normal adolescents during puberty. ¹ They used oral glucose tolerance testing and an euglycemic-hyperinsulinemic clamp in pre-pubertal and pubertal children without diabetes to investigate pubertal increases in insulin requirements.

Irrespective of glucose dose, glucose testing yielded a threefold greater insulin response to euglycemia in pubertal compared with prepubertal children. The clamp procedure showed that the prepubertal children were approximately 30% more sensitive to insulin than their pubertal counterparts. Interestingly, prepubertal girls were already more resistant to insulin than boys, presumably because of increased adiposity. (Figure 2).

Figure 2: Changes in Insulin Sensitivity During Puberty: Assessed Via Euglycemic Clamp

Hence, puberty is associated with a decreased sensitivity to insulin, which normally is compensated for by increased insulin secretion. Increased pulse amplitude of GH likely induces the insulin resistance that occurs during puberty.

Caprio and colleagues also found an increased insulin secretion during puberty, to compensate for reductions in insulin sensitivity. ² They examined the response to a standard hyperglycemic clamp in healthy preadolescent and adolescent children and in adults. A biphasic response was observed in preadolescents, with the first phase likely representing the preformed insulin already aligned on the microtubular microfilamentous system. In adolescents, a threefold increase in insulin was observed with both the first-phase and second-phase response. The adult secretion rate of insulin was near that of preadolescents. Despite the sharp increases in insulin secretion in the adolescents, the amount of glucose needed to maintain hyperglycemia was similar in all three groups. The differences found only in the adolescent group suggest that GH is responsible for the decreased insulin sensitivity and increased insulin secretion observed during puberty.

A longitudinal study of insulin resistance and metabolic changes during puberty came to a similar conclusion. ³ The study used hyperinsulinemic-euglycemic and hyperglycemic clamp studies in a cohort of prepubertal patients and then in the same patients during puberty. They found that insulin sensitivity diminishes by about 50% during puberty and is compensated by a doubling in insulin secretion. (Figure 3).

Figure 3: Longitudinal Study of Insulin Sensitivity

Changes in Protein and Fat Metabolism

Rather than resistance to insulin as found for carbohydrates, the effects of insulin on protein metabolism are greater in the presence of increased GH. During hyperglycemic clamp, Caprio and colleagues examined the fall in branched chain amino acids leucine, isoleucine, and valine in preadolescents and adolescents. ² At 90 and 120 minutes, the reduction in amino acids in the adolescents was much greater (by as much as twofold), revealing the effects of insulin extracting amino acids from the circulation and incorporating them into tissue.

Mauras and colleagues examined the effects of insulin-like growth factor I (IGF-1) and GH treatment in severely GH-deficient young adults on whole body protein synthesis rates. ⁴ Treatment with IGF-1 had a modest impact on protein synthesis in patients who were GH receptor-deficient. In patients who were GH-deficient, treatment with recombinant human GH restored whole body protein synthesis to the rate seen in the normal population.

A longitudinal study by Hannon and colleagues that compared prepubertal and pubertal stages found that turnover of glycerol is markedly increased during puberty, indicating a higher rate of total body lipolysis.³ The rate of oxidation of fat, expressed as the ratio of fat oxidation to glucose oxidation, also markedly increases. The body preferentially burns fat during puberty, and this study found that fasting free fatty acids actually decline. (Figure 4).

Figure 4: Whole Body Protein Synthesis Rates

Metabolic Changes with Growth Hormone Therapy

In adolescent boys with idiopathic short stature, 4 months of treatment with GH affects body composition, hormonal profile, and fasting lipid profile. ⁵ The non-GH-deficient children were studied in this way to simulate the pubertal increase in GH secretion. IGF-1 and muscle mass increased with GH treatment, while fat mass and percentage of body fat decline significantly. GH produced profound effects on the fasting lipid profile, with declines in total cholesterol and low-density lipoprotein cholesterol. The level of triglycerides increased, but free fatty acids declined. The study demonstrates that in the presence of GH, more fat is burned and more amino acids are extracted to create muscle. More insulin is secreted, which synergizes with GH to enable this process.

Pubertal Increases in Sex Steroid Hormones and Growth Hormone

Along with the increases in GH and insulin that occurs during puberty, there is increased secretion of the sex steroids testosterone and estrogen. Testosterone has a powerful effect on protein anabolism and lipolysis in the presence of GH. In adolescent boys, fat free mass increases, the percentage of fat mass declines, and linear growth accelerates.

Estrogen in the presence of GH increases protein anabolism, but to a lesser degree than does testosterone. Estrogen also restrains the rate of lipolysis and lipid oxidation. In girls during puberty, fat free mass increases slightly, the percentage of fat mass increases, and linear growth occurs. Fat free mass increases in both pubertal boys and girls, but less so in girls. Hence, GH synergizes with insulin and sex steroids to differentially affect boys and girls.

Glucose Intolerance During Growth Hormone Therapy

One should be aware of the possibility of untoward metabolic side effects of GH therapy, especially in children who are treated with large doses. It can unmask impaired glucose tolerance or even diabetes in children who have some impaired insulin secretion mechanism, either genetic or acquired.

Researchers from New Zealand undertook a retrospective analysis of data from an international pharmacoepidemiologic survey of children treated with GH (KIGS database) to determine the incidence of impaired glucose tolerance and diabetes. ⁶ Of the 22,333 children who were included for analysis, 85 were reported to have abnormal glucose tolerance, with 43 confirmed. Of the 43, 11 had type 1 diabetes, 18 had type 2 diabetes, and 14 had impaired glucose intolerance.

The incidence of type 1 diabetes did not differ from what was expected in the population, but the incidence of type 2 diabetes (34.4/100,000 GH treatment years) was sixfold greater than that in untreated children. Type 2 diabetes did not resolve in these children after discontinuing GH treatment. The authors of the study postulate that the higher than expected incidence of type 2 diabetes with GH therapy may indicate an acceleration of the disorder in predisposed people.

Conclusion

During puberty, GH works synergistically with insulin, estrogen, and testosterone for growth, development, and maturation of the human body. Increased physiological GH secretion in puberty leads to transient insulin resistance and increased insulin secretion. The increased insulin synergizes with growth hormone to enhance protein anabolism, and sex steroids further enhance this effect, testosterone more so than estrogen. GH enhances lipolysis and fat oxidation, sparing glucose and amino acids for anabolism and growth.

People who cannot adequately compensate for the pubertal induced insulin resistance by secreting more insulin can develop varying degrees of glucose intolerance. This development of glucose intolerance can occur in those genetically predisposed to having an impaired insulin response, in those with pancreatic autoimmune disorders or in those whose with risk factors for glucose intolerance such as obesity. If the body is already producing as much insulin as it can, such a child undergoing puberty or receiving GH therapy can develop type 1 or type 2 diabetes.

Similar processes can also occur during pregnancy; gestational diabetes can be seen as an outcome of the processes that occur during puberty. These processes explain the increased requirement for insulin during puberty and pregnancy.

References

1. Bloch C, Clemons P, Sperling M. J Pediatr. 1987;110:481-487.
2. Caprio S, Plewe G, Diamond M, et al. J Pediatr. 1989;114:963-967.
3. Hannon T, Janosky J, Arslanian S. J Pediatr. 2006;60: 759-763.
4. Mauras N, O’Brien KO, Welch S, Rini A, Helgeson K, Vieira NE, Yergey AL. J Clin Endocrinol Metab. 2000;85:1686-1694.
5. Hannon TS, Danadian K, Suprasongsin C, Arslanian SA. J Clin Endocrinol Metab. 2007;92:3033-3039.
6. Cutfield WS, Wilton P, Bennmarker H, Albertsson-Wikland K, Chatelain P, Ranke MB, Price DA. Lancet. 2000;355:610-613.

IGF-1-Based Growth Hormone Dosing for Treating Short Stature
Pinchas Cohen, MD

Growth is a result of both GH secretion and sensitivity, and in some cases, a problem with one or the other predominates. But in the large population of patients with idiopathic short stature, both GH secretion and sensitivity may be abnormal, making diagnosis and management especially difficult.¹

Clinicians are starting to move away from weight-based dose adjustment, recognizing that people vary in response to GH. Measuring IGF-1 allows us to select doses that might be more appropriate for each patient.

Genetic Differences to GH Responsiveness Identified

A number of studies have assessed GH response as a function of dose. A clear dose-response relationship has been found in GH deficiency. But for the more common patient with varying degrees of GH sensitivity, the picture is less clear.

Genetic abnormalities affecting the GH receptor have been found that control GH responsiveness, the best studied being the delta 3-GH receptor (d3-GHR) polymorphism.^2-13 For patients with this polymorphism, low doses of GH (30 µg/kg/day or less) have no effect on growth, but larger doses lead to better growth. Patients who are small for gestation age or who have acromegaly (who have elevated spontaneous GH secretion) respond better to GH if they have this polymorphism.

Additional clinical research is needed on the role of genetic abnormalities in short stature. Better assays are needed that go beyond identifying defects that cause severe growth disorders and that quantify responsiveness to GH treatment. Such assays will likely one day routinely help guide the evaluation and treatment of growth disorders.

IGF-Based Dose Titration to Guide Therapy

Insulin-like growth factor-1 (IGF-1) is well-recognized as a marker of GH action. Although IGF-based dose titration is controversial, it is part of the routine management for adults with GH deficiency and other endocrine disorders and could provide better efficacy and safety for pediatric patients.

A number of studies have shown that IGF-1 levels correlate with GH treatment response. A high dosage of GH (100 µg/kg/day) results in augmented growth and a large variation in individual IGF-1 levels: about half of patients develop IGF levels outside the upper limit of normal, and 25% above 3 standard deviations (SD).^14,15 Extreme elevations in IGF-1 for long periods of time should be avoided in clinical practice because of safety concerns (Figure 1).

Figure 1: Dose-Response and Variability in Outcomes of GH Rx

Whether IGF-based dose titration for guiding GH treatment in children should be routinely done is controversial. Even more controversial is whether titrating to a certain IGF-1 level is relevant to efficacy, but emerging data support that it is.

A linear relationship exists between GH dosage and growth in GH-deficient patients as expressed by either change in height standard deviation score (SDS) or the rise in IGF-1 over 2 years. Even more striking than the modest but significant dose response is the dramatic variability in response to dose, despite that in theory, GH-deficient patients are a homogeneous population: the change in height ranged from little more than .5 SD to as much as 3.5 SDs over 2 years.⁵

A recent study of IGF-based dosing evaluated whether titrating GH therapy to a specific target IGF-1 level correlates with growth.¹⁶ Patients were randomly assigned to GH doses titrated to achieve either the mean IGF-1 Z score, or the upper limit of normal Z scores or to a control group receiving 40 µg/kg/day.

The dosages of GH required to achieve mean IGF-1 varied from 20 to 100 µg/kg/day, with about half the patients requiring less than standard doses. Titrating GH to achieve the upper limit of IGF-1 to the upper limit of normal led to substantially more growth but the dosages required to maintain that level were substantially higher and were wide-ranging, up to more than 250 µg/kg/day (Figure 2).

Figure 2: Main Finding in 2051 (JCEM 2007) Delta Height SDS Over Time

Traditional models using single, low-dose approaches are based on patient age, gender, and weight, but these factors were not found to be predictive of the growth response. The peak GH dosage and the IGF-1 level were found important in determining the response to GH, as was the cumulative dose of GH, indicating that part of growth is IGF-independent and GH-dose dependent. Neither free IGF-1 nor IGF binding protein-3 predicted growth outcome.

These recent findings of the 2007 JCEM study prompted responses questioning whether this approach is ready for clinical practice.^16-18 Clearly, titrating GH to achieve IGF-1 levels at the upper limit of normal should not be done routinely, although it may be an option in select patients.

Most important, the study shows that IGF-based dose titration is a rational and feasible approach to managing GH treatment and overcomes the great variability seen in GH responsiveness. Ultimately, the target IGF level should be a clinical choice based on such factors as height, disease, and side effect risk. For example, titrating to -1 SDS or -2 SDS might be chosen if cancer is of concern in a susceptible patient.

Dose-Response Curves Should Guide Therapy

GH dose-response curves that are gender-specific across all ages and for multiple diagnoses have recently been published.^19-21 The curves incorporate the concept of GH response modeling that is very popular in Europe and include multiple parameters, such as peak dosage, age, gender, and weight. The model quite accurately predicts a patient’s height in 1 year with an expected variance.

These curves are a physician-friendly, elegant tool for identifying non-responders and enables physicians to select an optimal starting dose that is likely to result in the desired outcome.

Prediction models are currently being investigated in randomized trials. Hopefully, similar curves will be generated for patients treated with different GH dosages as well as for the IGF response.

Dosing Transition Patients

Dosing of adults with GH deficiency is another area that needs more study and for which IGF-based dosing may be useful. Currently no consensus exists on how to transition from pubertal dosing regimens, which can be as high as 50 to 100 µg/kg/day, to adult dosing, which may be one-fifth lower than that.²² Using well-defined ranges for IGF-1 during transition to adulthood could offer a convenient way to lower the dose in a physiological manner.

Similarly, transitioning from pre-pubertal to pubertal dosing could also be based on the age-specific IGF-1 norms rather than empirically doubling the dose when the first sign of puberty appears.

Summary

The traditional model of basing GH dosage on body weight does not account for individual differences in responsiveness. IGF-1-based dosing is a feasible approach to managing GH treatment and overcomes the great variability seen in GH responsiveness. The target IGF-1 level should be a clinical choice based on such factors as height, disease, and side effect risk.

References

1. Cohen P. J Clin Endocrinol Metab. 2006;91:4235-4236.
2. Lee KW, Cohen P. Horm Res. 2001;56 Suppl 1:29-34.
3. Dos Santos C, Essioux L, Teinturier C, Tauber M, Goffin V, Bougnères P. Nat Genet. 2004;36:720-724.
4. Blum WF, Machinis K, Shavrikova EP, et al. J Clin Endocrinol Metab. 2006;91:4171-4174.
5. Pilotta A, Mella P, Filisetti M, et al. J Clin Endocrinol Metab. 2006;91:1178-1180.
6. Binder G, Baur F, Schweizer R, Ranke MB. J Clin Endocrinol Metab. 2006;91:659-664.
7. Jorge AA, Marchisotti FG, Montenegro LR, Carvalho LR, Mendonca BB, Arnhold IJ. J Clin Endocrinol Metab. 2006;91:1076-1080.
8. Jensen RB, Vielwerth S, Larsen T, Greisen G, Leffers H, Juul A. J Clin Endocrinol Metab. 2007;92:2758-2763.
9. Schmid C, Krayenbuehl PA, Bernays RL, Zwimpfer C, Maly FE, Wiesli P. Clin Chem. 2007;53:1484-1488.
10. Tauber M, Ester W, Auriol F, et al. NESTEGG group. Clin Endocrinol (Oxf). 2007;67:457-461.
11. Carrascosa A, Audí L, Fernández-Cancio M, et al. J Clin Endocrinol Metab. 2008;93:764-770.
12. Carrascosa A, Audí L, Esteban C, et al. J Clin Endocrinol Metab. 2008;93:147-153.
13. Räz B, Janner M, Petkovic V, et al. J Clin Endocrinol Metab. 2008;93:974-980.
14. Tanaka T, Cohen P, Clayton PE, Laron Z, Hintz RL, Sizonenko PC. Growth Horm IGF Res. 2002;12:323-341.
15. Mauras N, Attie KM, Reiter EO, Saenger P, Baptista J. J Clin Endocrinol Metab. 2000;85:3653-3660.
16. Cohen P, Rogol AD, Howard CP, Bright GM, Kappelgaard AM, Rosenfeld RG; American Norditropin Study Group. J Clin Endocrinol Metab. 2007;92:2480-2486.
17. Baron J. J Clin Endocrinol Metab. 2007;92:2436-2438.
18. Chernausek S. Nat Clin Pract Endocrinol Metab. 2007;3:682-683.
19. Bakker B, Frane J, Anhalt H, Lippe B, Rosenfeld RG. J Clin Endocrinol Metab 2008;93:352-357.
20. Geffner ME, Dunger DB. . Horm Res 2007;68 Suppl 5:51-56.
21. Land C, Blum WF, Shavrikova E, Kloeckner K, Stabrey A, Schoenau E. . J Pediatr Endocrinol Metab 2007;20:685-693.
22. Ho KK, GH Deficiency Consensus Workshop Participants. Eur J Endocrinol. 2007;157:695-700.

Emerging Uses of Human Growth Hormone
David B. Allen, MD

Early Growth Hormone Therapy for Prader-Willi Syndrome

Prader-Willi syndrome was the first pediatric condition for which GH therapy was tried primarily to improve body composition; a secondary effect was improvement in neurodevelopmental milestones. Studies showed that although body composition improved, it did not normalize even after 4 years of treatment. Whether starting therapy at a younger age could confer greater benefit was still an open question.

Our baseline study of infants with Prader-Willi syndrome found fairly remarkable differences in body composition compared with healthy infants based on published normative data.¹ However, energy expenditure per gram of lean body mass followed the normal curve.

Further studies showed that infants and toddlers (ages 4 months to 37 months) with Prader-Willi syndrome treated with GH had decreased fat mass, increased lean body mass, and increased energy expenditure (measured with deuterium-labeled water) compared with untreated children of the same age with the syndrome.^2,3 Those who started treatment before 18 months of age acquired some developmental milestones significantly faster than the untreated children: they walked independently at 23 months vs 26 to 30 months and spoke their first word at 14 months vs 18 to 23 months. Children starting treatment who are older than 18 months of age did not differ significantly from untreated children in neurodevelopmental measures (Figures 1 and 2).

Figure 1: Effect of GH on Mobility and Neurodevelopment Scores in Very Young Children with PWS

Figure 2: HGH for CF: Effects on Height and Weight Velocity

More research is needed, but these studies strongly suggest that early treatment with GH is effective for children with Prader-Willi syndrome, many of whom have significant growth-hormone deficiency.

While concerns about increasing diabetes have not been realized, some safety questions remain. Specifically, reports of precipitous death in GH-treated children with PWS (who already have an increased risk of such events) have raised the question whether, still unresolved, of whether GH might be a contributing factor.

The optimal duration of treatment is also unclear. Whether therapy is sufficiently beneficial to continue therapy after adolescence, in the midst of sex hormone replacement (which is also used for some children with this syndrome) and during adulthood is still under investigation.

Growth Hormone Therapy for Some Other Conditions

GH therapy in children with certain chronic illnesses isprovided not only to increase height, but also for body composition and metabolic improvement.

Cystic Fibrosis

Hardin and colleagues randomized 61 prepubertal children with cystic fibrosis to daily GH or placebo for 1 year, then crossed over the groups for the second year of the study.⁴ Treated children had significantly greater gain in height, weight, lean mass, and bone mineral content, and had fewer hospitalizations. In addition, measure of quality-of-life measures of weight and body image. Somewhat disappointingly, however, was that pulmonary function did not differ between the two groups. A similar study of 63 patients in Germany that looked at short-term GH therapy (24 weeks) found improved pulmonary function in the treatment group, although the differences did not reach statistical significance. ⁵

These studies provide strong evidence for an anabolic effect of GH therapy in children with cystic fibrosis, some evidence that the children have improved quality of life and fewer hospitalizations, and a possible trend toward improved pulmonary function.

The pulmonary community tends to take only a sporadic interest in GH therapy for cystic fibrosis, and whether FDA approval of this therapy for the disease is likely to occur remains uncertain.

Juvenile Idiopathic Arthritis

Evidence of the benefits of GH therapy for juvenile idiopathic arthritis is stronger than for cystic fibrosis. Bechtold and colleagues randomized 31 growth-retarded children with juvenile idiopathic arthritis to weekly GH treatment for a mean of 6.7 years and found that long-term therapy improved growth and final height. Treated children had more bone mineral content, cortical bone, and muscle cross-sectional area compared with untreated children, but improved function was not shown.

No detrimental effects on disease activity, skeletal complications, or carbohydrate metabolism were observed in the treatment group. Benefits were greatest in children with moderate disease and moderate glucocorticoid requirements.

Growth Hormone as a Performance Enhancer in Athletes

In theory, the performance-enhancing effects of exogenous GH are many, based mostly on the effects observed when GH is restored to people who are GH-deficient. Benefits include increased metabolic function and improved body composition. Connective tissue repair may also be enhanced: collagen synthesis following injuries decreases with an age-associated reduction in GH production. Such benefits lead one to wonder: would more GH be better for a normal person? (Figure 3).

Figure 3: Theoretical Performance Enhancing Effects of Exogenous Growth Hormone

Providing GH to normal adults leads to some body composition changes, with men tending to have more of an increase in fat-free mass than women.⁷ Energy substrates are also changed: supplemental GH leads to increased lipolysis and fatty acid availability at rest and during and after exercise and to less utilization of protein as an energy source.^8,9 Interestingly, normally during exercise, people have a large spike in GH production, which gradually decreases over several hours, but people getting supplemental GH do not show such a pattern (Figure 4).

Figure 4: Body Composition Effects of One Month hGH in Young Active Healthy Adults

Evidence that GH Increases Performance is Lacking

Studies suggest that supplemental GH might offer some energy advantages. However, no study in the literature has thus far shown a performance benefit of GH therapy, either with regard to strength or endurance.^10-12

Hansen and colleagues studied well-trained male athletes during 120 minutes of aerobic exercise.¹³ Compared with placebo, a single dose of GH led to higher levels of non-esterified free fatty acids in the blood but no change in fat oxidation; so how the body actually used energy was unchanged.

Berggren and colleagues gave 4 weeks of GH or placebo to 30 active young people and could not document any benefit to several performance outcomes during bicycle exercise.¹⁴

The studies showing no benefit of GH administration to athletic performance is in contrast to anecdotal reports from coaches and in sports magazines that a definite benefit is observed. It is possible that scientific studies thus far have been inadequate because the most elite athletes have not been studied, or that GH was not administered for a sufficient duration to see benefits.

It is also possible that exogenous GH allows athletes to better tolerate the day-to-day strains of the sport rather than that improving performance measures. Given the actions of GH, it is certainly plausible that GH supplementation (particularly in an athlete in late 20’s or older) improves recovery from workouts or helps with injury prevention or the healing of minor injuries. These theories could be studied with surrogate markers, such as evaluating collagen responses to injuries or to heart training bouts.

Should GH use in Athletics Be Banned?

GH has been included in the same basket with anabolic steroids as a banned substance because of the assumption that it enhances performance. However, as described above, there is no scientific evidence at this point that GH is an effective performance enhancer.

It is also argued that exogenous GH should be banned on the basis that its use could be dangerous. But any risk of GH therapy pales in comparison to the risk assumed routinely by athletes: e.g. careening down the side of a mountain on the Tour de France bike race or bashing one’s head against the opponent’s on the line of scrimmage. Many elite athletes have already made the decision to lead very risky lives.

Although GH may cause some side effects, they are usually temporary, and concerns about long-term adverse events are largely theoretical, especially with episodic use. Recall that GH is considered sufficiently safe by the FDA for treatment of short, but otherwise healthy children without GH deficiency, and treatment that is supra-physiological in dosing and often persists for many years.

I suspect that athletes and trainers have observed real benefits that we have not yet been able to document. The interesting conundrum to ponder is: if it turns out that GH is an effective treatment that speeds the healing of sports-related injuries, then should its use be regarded as cheating or as medical therapy?

References

1. Bekx MT, Carrel AL, Shriver TC, Li Z, Allen DB. J Pediatr. 2003;143:372-376.
2. Carrel AL, Moerchen V, Myers SE, Bekx MT, Whitman BY, Allen DB. J Pediatr. 2004;145:744-749.
3. Myers SE, Whitman BY, Carrel AL, Moerchen V, Bekx MT, Allen DB. Am J Med Genet. 2007;143:443-448.
4. Hardin DS, Adams-Huet B, Brown D, et al. J Clin Endocrinol Metab. 2006;91:4925-4929.
5. Schnabel D, Grasemann C, Staab D, Wollmann H, Ratjen F. Pediatrics. 2007;119:e1230-e1238.
6. Bechtold S, Ripperger P, Dalla Pozza R, et al. J Clin Endocrinol Metab. 2007;92:3013-3018.
7. Ehrnborg C, Ellegård L, Bosaeus I, Bengtsson BA, Rosén T. Clin Endocrinol. (Oxf) 2005;62:449-457.
8. Healy ML, Gibney J, Russell-Jones DL, et al. J Clin Endocrinol Metab. 2003;88:5221-5226.
9. Healy ML, Gibney J, Pentecost C, et al. J Clin Endocrinol Metab. 2006;91:320-327.
10. Yarasheski KE, Zachweija JJ, Angelopoulos TJ, Bier DM. J Appl Physiol. 1993;74:3073-3076.
11. Yarasheski KE, Campbell JA, Smith K, Rennie MJ, Holloszy JO, Bier DM. Am J Physiol. 1992;262(3 Pt 1):E261-E267.
12. Deyssig R, Frisch H, Blum WF, Waldhör T. Acta Endocrinol. (Copenhagen) 1993;128:313-318.
13. Hansen M, Morthorst R, Larsson B, et al. J Physiol. 2005;567(Pt 3):1035-1045.
14. Berggren A, Ehrnborg C, Rosén T, Ellegård L, Bengtsson BA, Caidahl K. J Clin Endocrinol Metab. 2005;90:3268-3273.

Treatment of Growth Hormone Deficiency in the Transition of Adolescence to Adulthood
Nelly Mauras, MD

Although the average male finishes growing between 16.5 to 17 years, and females between 14.5 to 15 years, muscle mass and strength, as well as bone mineral density do not peak until the mid-twenties. Growth hormone (GH) is necessary for the continued changes in body composition and bone mineralization even past the years of completion of linear growth. This period when the child is done growing, yet has not achieved the peak muscle and bone development of adulthood has been termed the “transition phase”. For adolescents that were treated as children with GH, establishing the correct diagnosis of GH deficiency is therefore imperative so that patients who are truly GH-deficient can be selected for lifelong therapy while others are spared.

Up to 70% of idiopathic isolated GH-deficient children are no longer GH-deficient when retested at the completion of their linear growth. As such, retesting of the GH reserves is necessary before committing patients to GH therapy for life, with the exception of patients with multiple pituitary hormonal deficiencies, most of whom may not require re-testing.

Defining Persistent GH Deficiency

Growth hormone production rates more than double during human puberty, and GH production rates remain relatively robust until the late 20s. During the transition phase, GH is still being produced at higher rates that in the rest of adult years. As a result, consensus guidelines that define GH deficiency as peak GH response as <3 ng/mL may well be too stringent and do not fully apply to the definition of GH deficiency in the transition phase.

Mahgnie and colleagues found that re-evaluation of the GH response to an insulin tolerance test (ITT) with a peak GH response of <6.0µg/L differentiated GH deficiency (either isolated or multiple pituitary hormone deficiencies) in adolescents in the transition phase (ages 17 to 25 years) from age- and sex-matched controls. ¹ There was only one case whose insulin-like growth factor-1 (IGF-1) level was similar to the controls (6.1 µg/L). The authors concluded that a cut-off of <3 µg/L to define GH deficiency to ITT in the transition phase was too restrictive, suggesting instead it should be <6.0 µg/L.

The healthy controls in this study had consistently higher IGF-1 levels than the GH-treated subjects, but the latter had a high degree of variability in their IGF-1 levels, which supports the concept that very low IGF-1 levels are helpful in establishing a diagnosis of GH deficiency but are of lesser value as they approach normal levels.

Efficacy of GH Treatment in Transition Phase

In a small placebo-controlled study of 19 patients (aged 16 to 26 years) with GH deficiency (13 with idiopathic GH deficiency), who were treated with GH for at least 3 years and who had been off treatment for a variable period, Vahl and colleagues observed an increase in the percentage of fat mass in the group randomized to placebo of 3.8% from baseline (P=0.01), compared with a non-significant 1% increase in those continuing on treatment. ² However, there were no significant increases in lean body mass after 12 months in either group, and no significant differences between the groups in lipids, blood glucose levels, insulin concentrations, measures of muscle strength, and quality of life.

In a slightly younger cohort (ages 14 to 20 years) of 24 subjects with GH deficiency (defined as peak GH during an ITT after 7 days off therapy of < 3 ng/mL) that had been on GH for at least 5 years,³ Carroll et al found no significant difference in fat mass between those who were taken off GH and those who continued treatment for 1 year. Lean body mass increased by 2.5 kg from baseline in the GH group; there was no change in the controls. There was no significant difference between the groups in lipid levels; insulin sensitivity was better in the control group. Bone mineral density (BMD) increased by 6% in the GH group (P<0.001 from baseline) compared with 2.4% in the controls (P=NS), with no significant difference between the groups.⁴

Perhaps the most compelling data to support the benefit of GH in the transition phase of patients with GH deficiency comes from Underwood and colleagues in a cohort of 64 patients younger than 35 years who had been off GH for more than 5 years. ⁵ Persistent GH deficiency was defined as a peak GH to a clonidine/L-Dopa test of <5 ng/mL. The patients were randomized to placebo, low-dose GH (12.5 µg/kg/day) or high-dose GH (25 µg/kg/day) for 2 years. BMD increased after 2 years in the GH groups in a dose-dependent manner. Fat mass in the placebo group increased by 2.3 kg, declined by 0.7 kg in those randomized to low-dose GH, and declined by 3.7 kg in those randomized to high-dose GH. Lean body mass increased similarly in the two GH groups (13.4%) and increased by 3.7% in the placebo group. There were no differences observed between groups in quality of life, findings on echocardiography, insulin levels, glucose levels, and lipid concentrations.

Shalet and colleagues found increases in bone mineral density of 8.4% and 9.5% after 2 years in a study of 149 patients (mean age 19 years) with GH deficiency (80% multiple pituitary hormone deficiencies) who were randomized to low- or high-dose GH, respectively, compared with a 5.6% increase in placebo recipients (P=0.008). ⁶

We also conducted a study of 58 adolescents (mean age 15.8 ± 1.8 years) diagnosed with GH deficiency in childhood who were at the end of their treatment for linear growth. ⁷ To be eligible, they had to have been growing < 2cm/year and their epiphyseal bone fusion by bone age X ray complete. While still on GH, a battery of tests was performed, including metabolic parameters (lipids, glucose, insulin), hormone concentrations, dual energy X ray absorptiometry (DEXA), heart echocardiography, carotid intima-media thickness, tests of muscle strength, exercise tolerance, and quality of life. After these baseline studies, GH was discontinued and 4 weeks later an ITT was performed. Persistence of GH deficiency was defined as a peak response to ITT <5 ng/mL. Those considered persistently GH-deficient patients were randomized to GH (20 µg/kg/day) or placebo injections; 18 patients now testing as GH-sufficient served as controls. As expected, GH-deficient patients who continued GH had a smaller decline in IGF-1 from baseline than those randomized to placebo. However, changes in BMD Z scores were not significantly different between the groups (including controls) at 12 months and 24 months. There were no changes in lean body mass between the groups either. There were also no between-group differences in insulin, glucose concentrations, lipid profiles, insulin-like growth factor binding protein-3, muscle strength, left ventricular mass, function, carotid intima-media thickness, exercise tolerance, and quality of life. These data suggested that adolescents with idiopathic GH deficiency who are in good metabolic status at the time of epiphyseal fusion may safely discontinue GH for at least two years with careful follow up to decide if reinitiation of GH therapy is warranted⁷ (Figures 1 and 2).

Figure 1: % Change in IGF-1 From Baseline Visit Median with 1st and 3rd Quartiles

Figure 2: BMD Z Scores in Adolescents in Transition

Consensus Statement: Therapy of GH-deficient Adolescents in Transition

A consensus statement on the therapy of GH-deficient adolescents in transition was drafted by Clayton and colleagues with the input of several societies prior to the publication of our above-mentioned study.^7,8 For patients diagnosed with GH deficiency in childhood or at the end of growth and puberty, they proposed to discontinue GH for at least 1 month. Those adolescents with a high likelihood of severe GH deficiency (i.e., multiple pituitary problems, anatomical lesions) were recommended to have serum IGF-1 measurement. A diagnosis of severe GH deficiency would be confirmed based on very low levels of IGF-1 and GH therapy would be resumed. In with isolated GH deficiency and a low level of IGF-1, GH stimulation tests are recommended. If the peak response to ITT is < 5 ng/mL, GH should be restarted. The diagnosis should be reconsidered in adolescents with normal levels of IGF-1 and a normal peak GH response to stimuli according to the authors. In those adolescents with a low likelihood of severe GH deficiency, it is recommended to undertake a GH stimulation test and measure IGF-1 levels. If both the GH stimulation test and the IGF-1 level are low, they recommend resuming GH treatment. If both tests returned normal values, the recommendation was to discharge unless there is a risk of an evolving endocrinopathy. If the tests are discordant, they recommend follow-up without GH treatment.

In 2007, Radovick and DiVall offered a more expansive interpretation of the published data with respect to evaluating the GH status of transition patients. ⁹ Patients with a high probability of persistent GH deficiency (acquired congenital, multiple pituitary hormone deficiencies, or congenital GH deficiency with structural abnormalities or confirmed mutations) would be labeled as persistently GH-deficient. In those with a moderate probability of persistent GH deficiency (acquired GH deficiency from tumors, irradiation, or surgery, or idiopathic multiple pituitary hormonal deficiencies), the authors recommend discontinuingGH for 1 to 3 months. If they have a low IGF-1 level, they are considered persistently GH-deficient, and treatment should be continued. Insulin provocative testing or GHRH-arginine was recommended in adolescents with a normal IGF-1. With ITT, the authors defined GH deficiency as a peak GH <5.1 ng/mL; with GHRH arginine, it was defined as a peak GH <4.1 ng/mL. Those who have peaks < 10 ng/mL on the GHRH arginine are considered to have “probable persistent GH deficiency.” Those with peak GH <10 ng/mL on ITT or GHRH-arginine and at low probability of persistent GH deficiency were labeled “partial GH deficiency or normal.” In adolescent patients with GH deficiency, these authors suggest discussing with the patient the risks and benefits of GH continuation, and consider treatment with 12 µg/kg/day. A discussion of risks and benefits of GH continuation should also be undertaken in adolescents with isolated GH deficiency who were considered GH-deficient by retesting, and at high risk for metabolic consequences based on a low lean body mass and low BMD. Those at low risk for the metabolic consequences, however, should have their BMD, fat mass, lean body mass, and quality of life monitored.

Treatment Goals for the Transition

The somatic and biochemical targets during the transition period are BMD accrual and IGF-1 concentrations. Pediatric endocrinologists should establish the need for treatment, and whenever possible, collaboration with adult endocrinologists is desired.

What determines the need for continuing or stopping treatment? In the transition patient with multi-pituitary hormonal deficiency with a low IGF-1 concentration, there is no need to retest GH secretory status, and GH treatment can be continued at a lower dose. If these patients have a normal IGF-1 concentration, I recommend retesting.

In idiopathic isolated GH-deficient patients, I recommend retesting them all. If they have a normal GH (> 6 ng/mL to ITT), no treatment is required. Those who have a low GH BMD, and IGF-1 concentrations upon retesting should be considered for treatment.

In those with idiopathic isolated GH deficiency who have a low GH level upon retesting yet a normal DEXA and normal IGF-1 concentration, the data demonstrate no negative impact of treatment discontinuation for 2 years. These patients should be monitored annually for BMD, lipid levels, and IGF-1 concentrations.

Determinants for Restarting Therapy

Restarting therapy in those with idiopathic isolated GH deficiency in childhood should be considered in patients with persistently low GH responses to ITT who have a decrease in lumbar BMD >1 Z score over the next 2 to 3 years and the absolute Z score is also below normal range. Low IGF-1 concentration for their age and sex would also bolster the argument for re-initiation of treatment. The data from the three studies reviewed previously suggest that the restart dosage should be 12.5 µg/kg/day.

Conclusions

Profound childhood onset GH deficiency associated with multiple pituitary hormonal deficiencies or anatomic malformations or organic etiologies should prompt continuation GH in transition doses after linear growth is completed. However, based on the data thus far, many GH-deficient adolescents who are in good metabolic status at the time of discontinuation of GH treatment (i.e., favorable lean body mass, BMD, and normal IGF-1 levels) may be able to discontinue GH for at least 2 years. If after careful follow-up the phenotype of adult GH deficiency is fully identified, then the decision to resume GH therapy can be reconsidered. The timeframe when this will develop may well depend on the individual patient, the severity of the GH deficiency, his or her level of fitness, and even antecedent GH doses. The treatment for idiopathic GH deficiency in adolescents in transition needs to be individualized before life-long treatment is considered.

References

1. Maghnie M, Aimeretti G, Bellone S, et al. Eur J Endocrinol. 2005;152:589-596.
2. Vahl N, Juul A, JØrgensen JO, Orskov H, Skakkebaek NE, Christiansen JS. J Clin Endocrinol Metab. 2000;85:1874-1881.
3. Carroll PV, Drake WM, Maher KT, et al. J Clin Endocrinol Metab. 2004;89:3890-3895.
4. Drake WM, Carroll PV, Maher KT, et al. J Clin Endocrinol Metab. 2003;88:1658-1663.
5. Underwood LE, Attie KM, Baptista J; Genentech Collaborative Study Group. J Clin Endocrinol Metab. 2003;88:5273-5280.
6. Shalet SM, Shavrikova E, Cromer M, et al. J Clin Endocrinol Metab. 2003;88:4124-4129.
7. Mauras N, Pescovitz OH, Allada V, et al. J Clin Endocrinol Metab. 2005;90:3946-3955.
8. Clayton PE, Cuneo RC, Juul A, et al. Eur J Endocrinol. 2005;152:165-170.
9. Radovick S, DiVall S. J Clin Endocrinol Metab . 2007;92: 1195-1200.