7.2.6 Growth-promoting agents for nongrowth hormone-deficient short children

Growth-promoting agents refer to compounds, existing naturally or otherwise, which, when given to a short child as a medication, will accelerate growth velocity (GV), bringing the child’s height closer to or within normal range. The first objective, normalizing height during the growth phase, may not normalize adult height, which is the second objective. The ideal growth-promoting agent should normalize both. The mechanism behind these two objectives is not fully understood. It seems that when the growth plate cartilage chondrocytes multiply, they also differentiate then stop multiplying. The balance between multiplication and differentiation of growth plate chondrocytes is key to normalizing final height.

Investigation of medical conditions in children leading to short stature, on the one hand, and gigantism, on the other, has contributed to the understanding of mechanisms accelerating growth and increasing adult height. Untreated growth hormone deficiency (GHD) leads to marked short stature in childhood, delay in bone maturation and variable reduction in adult height. Growth hormone given to short growth hormone-deficient children has been the first agent shown to accelerate growth and improve adult height (1, 2, and see Chapter 7.2.5).

Short GH-sufficient children tend to secrete less growth hormone than tall children (3). In contrast excessive growth hormone secretion leads to acromegalo-gigantism (4, 5). These facts provide the rationale that exogenous growth hormone given in addition to that produced endogenously may increase final height.

Low dose androgen in boys and very low dose oestrogen in girls are able to increase GV slightly, but without beneficial effect on adult height since oestrogens accelerate bone maturation (6, 7). At higher dose, sex steroids may even reduce adult height. In contrast, one case describing gigantism in a man with an inactivating mutation of the oestrogen receptor (OR) gene indicated that oestrogen insensitivity leads to delay in bone age and a prolonged growth phase (6). Another case describing gigantism in a man with an aromatase gene defect added further evidence that lack of oestrogen resulted in marked delay in bone maturation, allowing prolonged growth (7). Aromatase is the enzyme that naturally transforms testosterone into oestradiol and androstenedione into oestrone. These key observations has led to the development of strategies for short children aiming at prolonging the growth phase by preventing growth plate cartilage fusion and increasing adult height. Two approaches were investigated, the first using aromatase inhibitors to block aromatase activity, the second blocking the OR with OR antagonists.

Growth hormone treatment has been the first growth-promoting agent used in short children without growth hormone deficiency. In such short GH-sufficient children clinical trials have demonstrated that growth hormone treatment accelerates GV and can normalize height including adult height provided it is used for long enough and at a sufficient dose. Beyond growth hormone itself, agents that increase growth hormone secretion, such as growth hormone-releasing peptides (GHRP), or agents that aim at preventing chondrocyte differentiation, such as OR or aromatase inhibitors, are used alone or in combination as growth-promoting therapies.

Somatic growth stops when growth plates are fused. Therefore, the capacity of growth-promoting agents to normalize height is dependent on the short child’s residual growth potential, directly linked to the age: the younger the better.

By definition, the short child presents during the growing phase with a height below –2 SD for an age-matched normal range (or below the third percentile in some countries). The short condition may either be transient or persistant leading to short adult height. Target height (TH) is the mean height calculated from parental heights and it represents a height or ‘target’ that a child should reach, according to his genetic background (8).

Careful diagnostic evaluation of the short child (see Chapter 7.2.4) is needed prior to treatment consideration. This should be performed by an experienced paediatric endocrinologist, since there are numerous causes causing short stature, a minority relying on specific positive diagnostic tests, the majority having yet to be identified, including many likely to have a genetic cause (9–11). Some diagnoses can be solved by a simple diagnostic test, such as a karyotype for Turner’s syndrome, but usually experience is required to recommend more sophisticated investigations, for instance, those identifying a molecular defect in the increasing number of genes involved in growth control.

Diagnostic investigation of a short child should include predicting adult height, one important step for treatment decision. Short stature children can be split into two groups: those without poor final height prediction (‘transient short stature’) and those with poor final height prediction (‘permanent short stature’). Predicting adult height and growth potential is therefore an important but difficult step in short stature assessment. Bone age (BA) is a widely used marker of residual growth potential, but has important limitations. BA also correlates with global body maturation and, in patients where BA is discordant from CA, puberty usually starts when BA rather than, chronological age (CA) reaches 10–11 years in girls, and 12.5–13.5 in boys. Short children very often display a delay in BA compared with CA. One year of BA delay results usually in one additional year of growth. BA is used to correct adult height prediction (12). There is however an important exception to this rule: short children born small for gestational age (SGA) (after intrauterine growth retardation) (13). In short SGA although BA is delayed, it is not a reliable marker for growth potential assessment nor for estimating age at pubertal onset (10, 11, 13). Therefore BA should not be used in growth prediction for short SGA since it overestimates adult height (13).

With the exception of those with constitutional delay in growth (14) and those GH-deficient short children who have been adequately treated with growth hormone who should both achieve normal adult heights, the majority of short children who present with a height below –2 SD during infancy or childhood will stay short as adults. This includes conditions such as children born small for gestational age (SGA), several syndromes including Turner’s, Noonan’s, Prader–Willi, other short conditions such as chronic renal failure (CRF), β-thalassaemia, the broad spectrum of skeletal dysplasia disorders including achondroplasia (FGF receptor 3 gene defect), hypochondroplasia (some with FGF receptor gene defects), Leri–Weill dysplasia (with a defect in the SHOX1 gene) and, finally, two conditions without defined aetiology, ‘idiopathic short stature’ (ISS) and ‘familial short stature’ (FSS). Once a diagnosis has been assigned, it is then important to define the management of this condition and personalize it to each child.

Most of short stature conditions outlined above have been the subject of clinical trials over the last 20 years, using growth-promoting substances, with growth hormone being the most commonly tested. Data from these trials has shown that some of these conditions lead to a formal treatment indication according to regulatory authorities, and are broadly used, while other indications stay within the domain of clinical research.

Despite this extensive experience of treatment, the debate remains whether short stature is a disease or not. If not, should the most severe short stature be considered a disability warranting treatment? Patients, parents, physicians, health authorities, and societies have different views on this issue. This has led to some differences in treatment criteria both between and within the limited number of growth hormone approved indications. This deserves to be kept in mind and additional studies on short-term effects including metabolic, psychological status and social adjustment, long-term outcome, as well as cost-benefit are needed. However, if shortness is defined as height below –2 SD, a –2.5 SD cut-off is considered to be of sufficiently severe short stature as to merit treatment for most indications. Such growth-promoting treatment should be recommended ‘for very short children whose ability to participate in basic activities of daily living is limited as a result of their stature’ (15).

When growth-promoting therapy is considered, treatment objectives should be carefully defined. For the two objectives of normalizing height during childhood and as an adult, three main factors need to be considered: the severity of the shortness, the age at start, and the target adult height. The shorter the child, the longer it takes to normalize height; therefore, the younger the child should be when treated, provided the investigator has the choice.

It is important for both the parents and the child to understand treatment modalities, that there is variability in growth response, potential risks, and for some conditions as yet limited knowledge of long-term outcomes. This requires detailed discussion prior to treatment. Identifying the impact on the child’s self-esteem of his shortness is part of this assessment, as well as defining treatment objectives and patient/parent expectations that may not match the potential outcomes from the treatment. Ongoing psychological assessment may be needed.

There are several compounds to consider for growth-promoting therapy that can be classified into two groups. The first group includes growth hormone, GHRP, and insulin-like growth factor 1(IGF-1). They have the capacity to directly accelerate height velocity (HV). The second group of compounds includes gonadotropin-releasing hormone analogues (GnRH-a) and aromatase inhibitors. The aim is to prolong the growth phase by delaying the oestrogen-dependent growth plate fusion; these agents should improve final height, but lack direct growth acceleration effects. Importantly, experience with these drugs is still limited and formal registration for a growth indication has not yet been achieved. They can only be applied when sex steroids are secreted, that is at pubertal onset or during puberty.

Information in this section refers specifically and exclusively to human recombinant DNA GH whose sequence is identical to the natural pituitary GH1 gene product. The first approval for recombinant human growth hormone in the mid-1980s applied to growth hormone deficiency. GHD represents a very small percentage of those children presenting with short stature (an estimated 1/3000 children). This is the only indication where growth hormone is used as a substitution therapy. Most other short children presenting with a height below –2 SD (an estimated 75–90/3000 children) do not have GHD and growth hormone treatment, given above a substitution dose, adds to the endogenous secretion to stimulate growth both through IGF-1 mediated and direct growth hormone action.

Other indications in non-growth hormone-deficient short children (also referred to as ‘GH-sufficient’) have now been registered in many countries worldwide. These indications include Turner’s syndrome, chronic renal insufficiency (also named Chronic Kidney Disease) before transplantation, short children born small for gestational age (SGA) (including short children born after ‘intrauterine growth retardation’), short children with Prader–Willi syndrome, short children with Noonan syndrome, and finally short children with Idiopathic short stature, approved in the USA only (13, 16, 17, 18, 19). In these indications, heterogeneity of response is observed, both in short-term height improvement and long-term benefit including final height, and thus optimization of treatment regimens is essential, as well as long-term for adverse events. The treatment recommendation in some registered indications may vary from country to country in term of age at start, severity of shortness, and growth hormone dose. These registered indications and trials in nonapproved indications are summarized in Table 7.2.6.1. Most studies have focused on the first 2–3 years of treatment, with or without control groups. Very few studies have been conducted regarding adult height within the frame of a prospective controlled trial. The level of evidence-based efficacy-benefit ratio is reasonable, but varies among indications and requires periodic reassessment.

The dose-effect relationship is significant in the first three years of treatment but attenuates with time (l) (Fig. 7.2.6.1). Beyond the first 3–4 years of growth hormone treatment the dose-effect relationship is weak. This is referred to as the waning effect of growth hormone. This is an important fact to take into consideration since it has efficacy, possibly safety, and certainly cost implications. Another important factor in growth hormone treatment is the great variability in response amplitude both within and across growth hormone indications. This variability seems to be independent of growth hormone dose (Fig. 7.2.6.2).

When considering adult height normalization, the longer the treatment duration and the higher the growth hormone dose, the greater is the improvement to adult height (20, 21). A better responsiveness in younger patients has also been observed across indications (22) (Fig. 7.2.6.3a). Finally, improvement of height SDS during puberty is limited, stressing the importance of height normalization prior to puberty.

The concept of ‘catch-up and maintenance’ should also to be considered for growth hormone treatment. Clinical trials have documented that growth hormone treatment in children without GHD should be delivered until adult height since discontinuation results in loss of the benefit gained (‘catch-down’) (23). Inducing catch-up aims at height normalization during the initial phase of treatment then maintenance aims at maintaining that initial benefit until final height.

Since the pattern of response to growth hormone treatment is complex, models of prediction of response have been developed identifying auxological and conventional biomarkers as predictors (Table 7.2.6.2) (21, 22, 24). Although the use of these models is still limited, they do allow prediction of up to 60% of the variability of response observed and provide useful information on an individual child. Importantly, these models in GH-sufficient short children indicate that the growth hormone dose is the best predictor of response during the first year. The models also show that the magnitude of the first year response is the best predictor for the following years (24). Since the growth hormone dose is the only factor on which the physician may act, it is very important that response is good during the first year of treatment. Pharmacogenomic research should provide an opportunity to optimize treatment by identifying the genetic signature for an individual’s response to treatment. A normal variant within the growth hormone receptor gene in which the exon 3 is deleted has been reported to be associated in various growth hormone indications including SGA, GHD, and Turner’s syndrome with improved response to growth hormone (25, 26). Identifying other genetic markers that are associated with high or low growth response to growth hormone should improve prediction of response in an individual and potentially will bring new insights into pathways involved in growth and metabolic effects (27).

In summary, considering the pattern of response to growth hormone treatment in short nonGH-deficient children, it is important to start at the most efficacious safe dose, and at an age young enough to benefit long-term growth. When deciding when to start and at what growth hormone dose, it is therefore key to integrate how much height gain is needed to reach the normal range and for how long treatment needs to be given to achieve normal adult height. In some cases, where the short stature is less severe, good results may be achieved with lower initial growth hormone doses (28).

Short stature occurs in 8–13% of newborns born SGA, a condition defined by birth length and/or birth weight score (SDS) below –2 when compared with reference values for gestational age (13). The term IUGR refers to reduced fetal growth velocity; this can be associated with a birth size within normal range, although the majority will be born SGA. Growth hormone treatment trials in short SGA children have shown that growth hormone treatment at a dose ranging from 0.033 to 0.065 mg/kg per day for 5–7 years normalizes height during childhood (29, 30, 31) (Fig. 7.2.6.1).

This normalization includes head circumference (32). Final height may also be significantly increased, including those starting treatment at pubertal onset as shown in a randomized controlled prospective study (20, 22, 23) (Fig. 7.2.6.4). Target height was normalized in up to 85% in a Ducth study comparing two dose of 1 versus 2 mg/m² per day, including 98% with a final height within target height range (24). There is a modest effect of dose on final height (28).

There are additional benefit from growth hormone treatment in short SGA children, whose overall performance levels are less than average, albeit within normal range (43); IQ, behaviour, and self-perception improved over time in SGA adolescents from scores below average to scores comparable with their peers (34). In addition, children who achieved a height closer to that of their peers showed less problematic behaviour (34).

Growth hormone treatment can also normalize BMI in these short SGA children, with a trend towards normalization of lean body mass and reducing tissue fat mass, a situation that should not increase the risk for metabolic syndrome in the long term (31, 35, 37–40).

Following initial reports from the Barker group, several studies have confirmed that, among newborns born at term, those with relatively lower birth weights, including both AGA and SGA premature babies have a higher relative risk of health problems in adulthood, including associations with hypertension, obesity and type 2 diabetes now referred to as the ‘metabolic syndrome’ (36, 37–40). The postulated mechanism is that of a ‘thrifty phenotype,’ implying that intrauterine malnutrition leads the fetus to adapt in an attempt to maintain an adequate nutrient supply. During growth hormone treatment, changes in insulin: glucose ratios have been reported, but in most SGA patients, impaired glucose tolerance is transient and, in the vast majority of patients, glucose tolerance is not altered and the insulin/glucose ratio returns to pre-treatment values after growth hormone discontinuation (41, 42) (Fig. 7.2.6.5). Glucose intolerance and type-2 diabetes have been shown to be rare during growth hormone therapy (41). It seems safe to monitor glucose homeostasis during the course of growth hormone therapy in short growth hormone-sufficient SGA especially when BMI SDS is above the mean for age and or if family history is positive for obesity or type 2 diabetes (see section Safety of growth hormone therapy).

Turner’s syndrome was the first growth hormone-sufficient short stature condition approved for growth hormone treatment in many countries following extensive clinical trials with growth hormone. (17, 45, 46). Heights of girls with TS should be plotted on TS-specific growth curves (47). These studies shown convincingly that growth hormone can both accelerate growth and improve final height. The growth hormone dose range tested varied from 0.045–0.1 mg/kg per day.

Growth response seems not to be dependent on karyotype (48). Since the great majority of Turner’s girls do not progress spontaneously through puberty, limited data are available on the long-term effect of growth hormone alone. The practice prior to growth hormone treatment of giving the anabolic steroid oxandrolone to enhance growth has led to its use in combination with growth hormone therapy. A key randomized study used growth hormone treatment with or without oxandrolone (17). A subset of the growth hormone plus oxandrolone group reached a mean adult height of 152.1 cm compared with 150.4 cm in those receiving growth hormone alone (17). The estimated projected adult height improvement compared with historical untreated Turner’s girls was 8.4 cm in the growth hormone alone compared with 10.3 cm in the growth hormone plus oxandrolone group (17). In an oestrogen free group treated with growth hormone (0.10 mg/kg/day) from a mean age of 10.2 years with a mean duration of 5.1 years, the estimated gain in adult height was + 10.6 cm compared with the projected adult height using Lyon Turner growth reference (20, 45). Sas et al. reported a mean height improvement of +16 cm compared with the adjusted Lyon projection using a growth hormone dose of 0.09 mg/kg per day during a mean 4.8 years, but starting at a mean age of 8.1 years (31). Even though comparison is difficult between these two studies with different familial and genetic backgrounds, they both confirm that height can be improved dramatically and support that the younger the onset of growth hormone treatment the better the outcome.

In a large series from an international observational data based on more than 880 Turner cases, the estimated final height gain is +1 SD compared with Ranke’s Turner standards (46, 49). In the absence of placebo-controlled studies to final height, whose feasibility is questionable, these results strongly support the fact that growth hormone treatment has the capacity to both accelerate height velocity and improve adult height.

The influence of oestrogens deserves consideration. Standard substitution doses of oestrogens in these oestrogen deficient girls carry the risk of excessive bone maturation with loss of adult growth potential (50). Even very low dose oestrogens seem not to be beneficial to growth outcome (51). If growth hormone treatment has been started at a young age with good improvement in height, then low dose oestrogen can be started at the normal pubertal age (11–12 years) followed by progressive dose increases; this has the psychological advantage of not having to delay pubertal induction in these adolescent girls (52).

Prediction models of response to growth hormone treatment in Turner’s syndrome have been developed (21). They indicate that the best predictor of response in the first year of treatment is growth hormone dose, followed by age at onset of treatment, distance to target height and if included in treatment oxandrolone. In addition, the growth hormone receptor d3 polymorphism seems to be associated with a better growth response, indicating an important role for genetic factors (52).

Lack of controlled studies makes it difficult to assess the degree of quality of life improvement. Insulin-glucose homeostasis surveillance is also required in Turner’s syndrome (53). In addition, from a cardiovascular perspective, dilatation of the ascending aorta only seen on thoracic magnetic resonance imaging has been reported (55). Growth hormone treatment does not seem to alter this anatomical defect when present, but systematic surveillance prior to and after 1 or 2 years of growth hormone treatment is required. Oedema, but not arthralgia or myalgia have been more frequently reported in Turner’s syndrome than in other growth hormone-treated groups, as were slipped capital femoral epiphysis and scoliosis (56). Diseases associated with Turner’s syndrome, including thyroid disease, seem not to be influenced by the course of growth hormone treatment and only merit standard clinical surveillance (56). Overall based on large scale use over 20 years, growth hormone treatment is considered to be a safe therapy in Turner’s syndrome (57).

In summary, early diagnosis of Turner’s syndrome will allow the physician to develop a good interaction with the parents and the child and to start growth hormone treatment early. The use of Turner’s growth charts allows the projection of height in childhood and as an adult, helping to establish growth hormone treatment objectives. Growth hormone treatment should be initiated early (between 4 and 5 years of age) at a starting dose of 50 µg/kg per day, with response being assessed against that calculated from a prediction model. Monitoring of serum IGF-1 levels every 6 months helps with adjusting the growth hormone dose, while avoiding excessively high IGF-1 values, and should be targeted at IGF-1 values between +1.5 and +2.0 SDS. Oxandrolone should not be used at doses above 0.05 mg/kg per day and should not be given under 8 years of age. Bone age monitoring should also be undertaken. Low dose oestrogens (one-sixth to one-quarter of the adult dose) may then be introduced between by 12 years for 12–18 months with progressive increases in dose up to low substitution doses with routine assessment of uterine size on pelvic ultrasonography. Under these conditions, an adult height in the lower end of the normal range can be secured in the great majority of patients. A multidisciplinary team is beneficial and psychological support is often needed including transition to adulthood. Long-term surveillance including glucose metabolism is required (47).

The SHOX gene (short stature homeobox-containing gene) is located in the pseudo-autosomal region of distal short arms of the X and Y chromosomes. Mutations of SHOX generate a spectrum of phenotypes with short stature and skeletal dysplasia of various severity, including Leri–Weill dyschondrosteosis, Turner’s syndrome, and Langer’s mesomelic dwarfism (57). SHOX gene defects also account for ∼2% of so called ‘idiopathic short stature’ with a familial component. Growth hormone treatment in SHOX gene defects is approved in several countries; its management, efficacy and safety profile mimics that of Turner’s syndrome without oxandrolone treatment. In the absence of gonadal dysfunction the use of sex steroid supplementation is irrelevant (58).

The first publication to report that growth hormone treatment increased adult height in peripubertal children with idiopathic ISS was in 2004 (18). Growth hormone at a dose of 31 μg/kg per day or placebo were given in three doses per week starting between 9 and 16 years, and continued for 3–4 years. Reported adult height was greater in the growth hormone-treated than in the placebo-treated group by 0.5 SDS (3.7 cm). This initial study showed a significant, but limited gain, in part explained by the fact that both the growth hormone dose and frequency of administration were low. In another dose–response study, short ISS patients who received 53 μg/kg per day of growth hormone had a greater increase in height SDS than those who received 0.24 μg/kg per day—27% after 4 years of treatment and 42% in a limited number of subjects followed to adult height, respectively (59).

This indication is registered by the Federal Drug Administration in the US (2003), but not by the European Medicines Agency (2009). The reasons are complex. ISS is by definition a diagnosis of exclusion, but does account for a significant percentage of short children and, therefore, if all were treated with growth hormone a large cost would be incurred. In addition, in contrast to other short stature conditions with a medical aetiology, many authorities still perceive ISS as a short normal condition. There is an ongoing debate about whether shortness by itself is a pathological condition or not; however, the paediatric community considers that severely short children should be offered the opportunity of growth-promoting treatment (60). There are also limited data showing that normalizing height improves other features associated with shortness including low self-esteem and psychological issues. Quality of life studies are difficult in paediatrics due to the constant evolution of a growing child. In addition, change in height takes time. Trials powered for such end-points are very difficult to perform and would be needed to be placebo-controlled. Finally, the benefit of improving psychological status alone in children, although very important has limitations. Therefore, a decision about treating an ISS child requires evaluation of the estimated benefit risk ratio both in terms of height, psychosocial status, and the environment in which the child is being raised.

Recent studies point to the efficacy of recombinant human (rh)IGF-1 treatment in ISS children presenting with normal growth hormone levels, but with low IGF-1 (primary IGF-1 deficiency) (61). This emphasizes the need for carrying out controlled clinical trials of growth hormone vs IGF-1 vs combination therapy in this patient group.

Growth hormone treatment for Prader–Willi syndrome (PWS) is approved both by the EMEA and FDA authorities with a recommended growth hormone dose of 35 μg/kg per day. EMEA approval for PWS includes both growth failure and obesity. A height gain close to +2 SD was observed after 5 years of therapy (19). In addition a trend towards normalization of body mass index, body composition and improvement in both hypotonia and muscle function was observed. However a serious adverse event has been noted in this group with several deaths reported in very young, severely obese PWS patients with upper airway obstruction and central abnormalities in the control of respiration (62). These deaths tended to occur during the first year of growth hormone therapy. This has resulted in both a limited contraindication for some PWS patients and specific guidelines for identification of excessive obesity, upper airway obstruction and defects in respiratory control that should be managed specifically prior to growth hormone treatment.

After placebo-controlled trials, growth hormone treatment has been approved in many countries for patients with Chronic Renal Failure (CRF) (also referred to as Chronic Kidney Disease, CKD) prior to kidney transplantation (63, 64). Responses to growth hormone in doses ranging from 30 to 50 μg/kg per day is variable, but on average induces a height gain in the order of +2 SD after 5 years, indicating that the effect of growth hormone in this indication is ∼20–30% less than in GHD. Growth hormone treatment is also used after renal transplantation, but long-term safety in this situation requires additional documentation (65).

The mutation of PTPN11 gene encoding for SHP2, a nonreceptor protein tyrosine phosphatase, is found in half of all patients with Noonan syndrome. The SHP2 protein is involved in the down regulation of the GH receptor gene. Growth hormone treatment of Noonan’s syndrome has been developed through open uncontrolled limited scale trials, pointing to an apparent efficacy that mimics that observed in Turner’s syndrome with growth hormone alone (66), with growth hormone used in the dose range of 33–66 μg/kg per day. Adult height data ranging from 157.7 to 174.5 cm have been reported on a limited series with an estimated improvement of +9.8 cm in females and +13 cm in males (67). Although growth hormone treatment is approved by the FDA and in some countries like Switzerland, additional documentation on long-term efficacy and safety is needed.

One mutation of the FGF receptor 3 gene accounts for 98% of cases of achondroplasia, while several different mutations account for hypochondroplasia. Japan is the only country where growth hormone use is registered for achondroplasia (68). Growth hormone use in these skeletal dysplasias has shown limited and variable effects, most of the height improvement comes mainly from spinal growth (58, 69). The final height and long-term benefit is not yet established.

Growth hormone treatment has been tested in a long list of rare diseases presenting either with short stature and/or short predicted final height (below –2.0 SD), where growth hormone secretion is sufficient. Most of these conditions will never qualify for drug registration, as randomized placebo-controlled trials justifying their use are very difficult to perform. However, large international databases collecting in excess of 60 000 children treated over the last 20 years can provide a reasonable overview of experience with growth hormone in rare disorders (70). Such conditions include Silver–Russell syndrome, Down’s syndrome, congenital adrenal hyperplasia, cystic fibrosis, various skeletal dysplasias, a series of rare syndromes with short stature and rare medical conditions requiring glucocorticoid therapy and precocious puberty (71). The range of growth hormone doses used ranges from 33 to 65 μg/kg per day, matching that used in GH-sufficient approved indications. Despite these deficiencies, interesting preliminary information on short-term safety and efficacy of growth hormone treatment has been obtained, and ongoing efforts to collect data for evidence-based medicine should be pursued.

rhGH has been used for more than 25 years and data from more than 100 000 treated children has been collected in several large databases, including KIGS and NCGS, establishing a safe profile within the currently approved indication (55, 72). There are, however, some specific side effects. These include rare episodes of benign intracranial hypertension (also called pseudotumour cerebri) that tend to occur in the first phase of growth hormone use. The practice of starting growth hormone treatment at a lower dose, and increasing in a stepwise manner to the target dose over 4–6 weeks has been advocated, but has not been evaluated for its potential benefit in reducing prevalence of this side effect. Persistent headaches deserve evaluation, including ophthalmoscopy with transient discontinuation of treatment should papilloedema be present. Slipped capital femoral epiphysis is observed in hypothyroidism, GHD, and after body irradiation. Analysis from KIGS and NCGS databases does not support a causative role of growth hormone therapy. Analysis also does not support a role for growth hormone treatment in increasing the risk for diabetes mellitus (41). Although short SGA children do have an increased risk for the metabolic syndrome as adults, growth hormone does appear to be a safe treatment in SGA, as well as all GH-sufficient treated children, even when BMI SDS is above the mean for age and/or there is a family history of obesity or type 2 diabetes. There is, however, no consensus on how best to identify insulin resistance. Repeated random plasma glucose/insulin samples from which that a HOMA index can be generated or an oral glucose tolerance test (OGTT) seem more appropriate than monitoring haemoglobin A1c. Should insulin resistance or glucose intolerance be present, appropriate diet and life style should be advocated, as would be indicated for type 2 diabetes and/or obesity, and the growth hormone dose should be adjusted to a lower dose in order to minimize the growth hormone effect on insulin resistance without compromising the growth effect. Many other side effects have been reported during the many years that growth hormone treatment has been used. In most cases, there is no evidence of growth hormone treatment causality, but ongoing surveillance using international collaborative databases is required.

Growth hormone treatment is used in some patients after a first malignancy (where risk for relapse exists and risk for a second malignancy is greater than the average), and in patients carrying a known increased risk for malignancy (i.e. Bloom’s syndrome, Fanconi’s anaemia or Down’s syndrome) or an unknown carrier of a genetic cancer risk. It is very difficult to exclude the contribution of a treatment to malignancy risk unless a positive association starts to be observed. Growth hormone treatment has benefited from extensive and repeated analysis over the last 20 years using the two large databases (55, 72). With respect to growth hormone treatment and leukaemia, the only fact observed is that GH-treated patients who develop leukaemia do so at a mean age greater than the control population. Growth hormone treatment seems not to be associated with an increased risk of developing intracranial neoplasm with a possible exception for meningioma in GH deficiency treated after primary leukemia or lymhoma (Ref.: Ergun-Longmire B, A C Mertens, P Mitby, J Qin, G Heller, W Shi, Y Yasui, L Robinson & Ch A Sklar; Growth Hormone tretament and risk of second neoplasia in the childhood cancer survivor. Journal of Clinical Endocrinology & Metabolism 2006; 91: 3494–3498). Data on the risk of recurrence of central nervous system neoplasms with growth hormone treatment are reassuring, but not conclusive and deserve further follow-up and still larger patient numbers. There is a slight increase in second intracranial neoplasms, essentially meningioma, with link between growth hormone and irradiation therapy being unclear. As long as some uncertainty remains (‘we do not see but cannot exclude’), this is a difficult and sensitive domain when it comes to counselling parents and patients. The ‘no risk’ statement is not demonstrated and should not be used. The fact that cancer cell progression of existing neoplasms will likely benefit from new drugs opposing growth hormone-IGF-1 actions will increase this difficulty. However, currently growth hormone treatment in children seems not to be associated with de novo neoplasm and there seems to be no trend when comparing data at 5, 10, or 20 years (55). It seems appropriate to tell parents that, although data are reassuring, we need longer observation on larger patient number before definitive conclusions can be made.

Recombinant human IGF-1 became available in the 1990s for human treatment. It should be used for what is referred to as primary IGF-1 deficiency as opposed to IGF-1 deficiency secondary to GHD or malnutrition.

Growth hormone insensitivity syndrome is rare including its most recognized form, Laron’s syndrome, which is due to mutations within the GH receptor gene (GHR) (73). Insulin-like growth factor I (IGF-1) is now approved to treat this severe short stature disorder. The results of four different patient cohorts with GHR gene defects (in Israel, USA, Ecuador, and Europe) treated with IGF-1 are consistent with growth velocity increasing from a pretreatment mean of 2.5–5 cm/year to a first-year mean of 8–9 cm/year at doses ranging from 40 to 120 µg/kg twice daily by subcutaneous injection. In 67 growth hormone-insensitive patients, defined by failure to increase serum IGF-1 concentrations after four daily SC injections of growth hormone at a dose of 0.1 mg/kg, treated with IGF-1 for 1 year or more (74), growth velocity increased from 2.8 cm/year on average at baseline to 8.0 cm/year during the first year of treatment and was dependent on the dose administered.

Although IGF-1 induces catch-up growth as presently used, it does not reach the magnitude that is observed when severe growth hormone-deficient children are treated with growth hormone. The estimated benefit to final height is around 10 cm. The evolving experience with rhIGF-1 tends to confirm this observation, pointing to the fact that in severe growth hormone insensitivity syndrome, even though IGF-1 has the capacity to improve growth, it does not fully compensate for the effect of growth hormone (75). These observations suggest that growth hormone action on growth is not fully IGF-1 mediated and IGF-1 treatment lacks some direct effect of growth hormone. Specifically, growth hormone has important lipolytic effects reducing visceral fat in contrast to the IGF-1 lipotrophic effect, which is similar to insulin.

Practically, IGF-1 dose should be started at 40 µg/kg twice daily given subcutaneously and progressively increased to 80 up to 120 µg/kg per day three times daily over 4–6 weeks. The intravenous route is contra-indicated because of the risk of severe hypoglycaemia and hypokalaemia. Side effects include hypoglycaemia, requiring IGF-1 administration with/after meals, injection site lipohypertrophy and tonsillar/adenoidal hypertrophy.

Beyond growth hormone insensitivity syndrome, IGF-1 treatment could be used in very poor responders to growth hormone treatment within the ISS patient group. The IGF-1 generation test is useful to understand the level of sensitivity to growth hormone, but does not fully predict growth hormone responsiveness (76). Attempts to identify growth hormone unresponsiveness in children with ISS have yielded only a handful of patients with rare genetic disorders altering growth hormone receptor or postreceptor mechanisms or the development of growth hormone inactivating antibodies (77). Therefore, most poor responders to growth hormone present with functional resistance to growth hormone treatment caused by an as yet unidentified mechanism, but pointing to likely rare genetic causes. The conditions under which these poor responders to growth hormone therapy should be given a trial of IGF-1 have not yet reached a consensus. In such patients, the use of IGF-1 alone or combined with growth hormone is a matter for ongoing investigation.

IGF-1 treatment has also been used in very rare short stature conditions with IGF1 gene deletion (78), rare cases of Stat5b gene deletion, one of the growth hormone-receptor signal transduction molecules (61), and severe short stature with insulin resistance, Leprechaunism (79). These very rare conditions where IGF-1 treatment has been used by experienced academic teams require extended clinical research to improve documentation of the safety and efficacy. Finally, beyond growth-promoting effects, IGF-1 has many potential therapeutic uses because of its varied effects—insulin-like influence on glucose metabolism, and neuroprotection resulting from cell-proliferative, and anti-apoptotic properties. These have not been investigated systematically in clinical situations (80).

IGF-1 treatment should be prescribed according to the EMEA and FDA approved indications. IGF-1 for other indications should only be used within the frame of clinical research (81).

Growth Hormone-releasing hormone (GHRH) is the natural neuropeptide secreted by the hypothalamic arcuate and ventromedial nuclei that induces growth hormone synthesis and secretion by the pituitary somatotroph cells (82). GHRPs are synthetic peptides that bind to the growth hormone secretagogue receptor, whose natural ligand is Grehlin, a peptide secreted by the stomach that has two major effects, stimulating growth hormone secretion, and appetite/meal initiation (83). There is no registered use for GHRH or GHRP as growth or metabolism promoting drugs. However, there are several compounds that are being or have been tested for growth-promoting capacity including synthetic GHRH, GHRPs like GHRP2, including compounds given orally. Although these compounds carry the potential to be growth-promoting substances via induction of growth hormone secretion; however, their development remains speculative unless they provide either greater efficacy or easier route of administration or lower cost production compared with growth hormone (84, 85).

Aromatase inhibitors delay bone age thus prolonging the growth phase, carrying the potential of adult height improvement (87). Their specific use aims at improving final height and needs further development. If combined with growth hormone treatment when the residual growth potential is limited, they could lead to final height improvement.

Delaying puberty in short patients in the peripubertal period is feasible with GnRH antagonists (GnRH-a) commercially registered for precocious puberty. In the latter situation, they have shown benefit on final height improvement when given to suppress puberty starting prior to 7–8 years of age.

From these observations together with the known effect of aromatizable androgens or oestrogens on bone maturation, GnRH-a have been used either alone or combined with growth hormone therapy (88). Used alone in the peripubertal age (10–11 to 13–14 years in girls, 12–18 months later in boys) GnRH-a show no benefit on adult height and should, therefore, not be used. Results of combined GnRH-a and growth hormone therapy in the peripubertal age are contradictory, from no benefit to a significant gain in predicted adult height of 9.3 cm after 3 years of growth hormone + GnRH-a treatment, compared with a 1.2-cm gain in the control group (89). In some patient groups such as short SGA children, no benefit has been reported to recommend the combined use of growth hormone plus GnRH-a as routine treatment, considering the potential side effect on bone mineralization (89). Ongoing clinical trials should help to add more data on the appropriate use of GnRH-a together with growth hormone.

One agent that should be mentioned, although experience in the growth-promoting domain is very limited, is the insulin-sensitizer metformin. Many publications have discussed the potential role of insulin sensitivity on growth, based on basic observations that insulin regulates the expression of the IGF1 gene, and is key to glucose and amino acid delivery to growing tissues. Metformin has been extensively used for over 50 years in type 2 diabetes, is well tolerated and benefits from a limited paediatric indication in obesity. Since several short conditions are associated with some level of insulin resistance and, since exogenous growth hormone decreases insulin sensitivity, combined therapy with growth hormone and metformin has been shown to optimize total pubertal growth in short SGA girls (90). The potential interest in metformin beyond growth is that it might attenuate the effect of growth hormone therapy on insulin resistance, improving safety and efficacy at a low cost. Additional experience is required to establish the appropriate use of this combination in growth promotion.

A strategy for the use of costly drugs is essential. Treatment objectives should be defined carefully and shared with parents, including expectations on the magnitude of the catch-up phase and its duration, maintenance of initial benefit, the need to optimize treatment prior to puberty, and estimation of final height in relation to target. Out of approved indications, the use of growth hormone, IGF-1, or their combination, or combinations with other growth modulating agents should be carried within the framework of clinical trials aiming at clarifying predefined key objectives. Maintaining large global registries of treated patients collecting data on in particular long-term adverse events should be pursued, including post-treatment follow-up. The practice of individual IGF-1 titration in growth hormone therapy should be part of standard treatment and individual dose titration should be recommended (91). The use of prediction models, including new models benefiting from pharmacogenomic discoveries should also become part of standard clinical practice. Cost issues should always be in physician’s mind and the real benefit of height normalization deserves additional investigation. Poor responders remain an issue. Reassessment of treatment objectives after the first year (using prediction models) is needed. To date, there is limited experience of combining or switching growth-promoting agents and should be done with caution. In some cases, treatment discontinuation should be considered. Ultimately, the well being of short patient needs be kept in mind and psychological assessment and reassessment remains very important.