GH prescription

With the availability of biosynthetic GH there are undoubtedly more patients being treated than in the days of pituitary-extracted GH,14 although in practice, much of the GH treatment is still being instituted and run via the historic growth centres (approximately ∼80% of GH-treated children), compared to other tertiary centres or district general hospitals.

Details of existing14 and new15 paediatric patients on GH in the UK are shown in table 1.

On the basis of 760 new starts per year and 4712 patients on GH therapy, it would appear that paediatric patients in the UK are receiving GH therapy for an average of 6.2 years.

Indications for GH therapy

In addition to GHD, where the basis of therapy is physiological replacement, GH is now being used in non-GH deficient states associated with short stature, where the purpose of therapy is to normalise growth (height and also height velocity (HV)) both in the short and long term (ie, final height). In these situations GH is started either once the condition is first diagnosed, or once the HV falls below acceptable levels. In addition to improvement in height and HV, GH is also used in some indications (eg, Prader–Willi syndrome (PWS)) for improvement in body composition, along with other outcomes such as changes in bone density, cardiovascular risk, respiratory function, behaviour, socialisation and self-esteem.

GH is now licensed within the UK for the following indications, with the current percentage of GH-treated patients being14:

• ▶ GHD: 57.4%.

• ▶ Turner syndrome (TS): 18.7%.

• ▶ Small for gestational age (SGA): 5.2%.

• ▶ PWS: 4.6%.

• ▶ Chronic renal insufficiency (CRI): 2.5%.

• ▶ Short stature homeobox (SHOX) deficiency: unknown as licensed after 2006.

• ▶ Unlicensed indications: 11.6%.

The proportions of the different groups were effectively unchanged from a previous national survey in 1998, apart from a fall in unlicensed indications with the subsequent granting of product licenses for PWS, SGA and SHOX deficiency.16 The earlier paper did, however, show wide regional variation in prescription, with a twofold variation between regions for treatment of GHD. While this might merely reflect different diagnostic criteria there was in addition a more than threefold difference for TS.

In the USA, there is also a product license for idiopathic short stature and Noonan syndrome, and in Japan for chondrodysplasia.

NICE reviews

There have now been two reviews of GH in childhood by NICE; TA42 in May 2002,20 and TA188 in June 201021—the latter including the two GH licences granted after 2002 (SGA and SHOX deficiency). As with other reviews an evidence-based approach was taken, which in the case of paediatric GH therapy proved problematic, as it is considered first unethical both to do placebo controlled trials where daily injections are required and also not to treat licensed indications. As a result there is a paucity of randomised controlled trials (RCTs) involving either a placebo or untreated control group. Within the most recent review of 674 potential references, only 27 RCTs of varying quality were able to be included, which included at least one of the following outcomes: final height; height SD score (height SDS); growth velocity; growth velocity SDS; body composition; biochemical and metabolic markers; and adverse events. This includes only one RCT on GHD, two on final height (one TS and one SGA) and none on quality of life. No studies reported health-related quality of life. Overall the comments of NICE were that ‘the studies were generally poorly reported and some were of short duration’.21

GH deficiency

This is the commonest endocrine disorder presenting with short stature, it is estimated that a quarter of all children with a height below −3 SDS have GHD. Untreated final height is 134–146 cm in males, and 128–134 cm in females. The frequency of GHD is estimated at 1 in 3500–4000, although a milder phenotype may occur in up to 1 in 2000.20

Most (∼70%) of patients with GHD have an isolated deficiency of GH, but it can also occur as part of multiple pituitary hormone deficiencies.

The commonest causes of GHD are as follows:

• ▶ Congenital, due to midline embryonic defects and transcription factor mutations.

• ▶ Acquired, due to pituitary/hypothalamic tumour, trauma, irradiation, infiltration, infection.

• ▶ Idiopathic.

In many patients the cause is idiopathic. Temporary failure may also occur peri-pubertally, in association with emotional deprivation and also hypothyroidism. A national registry of GH treated patients showed that 22% of total GH-treated patients had idiopathic GHD, 8% congenital GHD, 5% craniopharyngioma and 18% other acquired GHD.22

The clinical features of GHD are dependent on:

• ▶ whether it is congenital or acquired

• ▶ whether it is isolated or associated with other multiple hormone deficits

• ▶ severity of hormone deficit.

The classical clinical phenotype, which is often present either at birth or within the first years of life, includes:

• ▶ short stature (both in relation to peers and also parents)

• ▶ poor growth (HV <25th centile for at least 1 year), and in severe GHD HV may be <4 cm/year

• ▶ delayed bone age (with associated delayed dentition and puberty).

In addition the following are noted:

• ▶ increased skinfolds

• ▶ small hands and feet

• ▶ ‘doll-like’ facies due to mid-facial hypoplasia

• ▶ micropenis in males

• ▶ hypoglycaemia may occur (especially in the neonate).

Milder forms of GHD may not be diagnosed until the child is older and other clinical manifestations at this stage are usually less obvious.

The GH units have recently changed, with results now expressed in μg/l, which produces levels three times lower than the previously used levels of mU/l. Prior to 1985 peak GH levels <2.3 μg/l (≡<7 mU/l) were considered to represent severe, and 2.3–5.0 µg/l (≡7–15 mU/l) partial GHD.5 NICE (http://www.nice.org.uk)20 now recommend that the diagnosis of GHD is confirmed by a peak plasma GH level <6.7 μg/l to two provocative tests, although for organic GHD only one provocative test may be required. There does remain, however, wide heterogeneity in diagnosis of this condition.23

Growth

In the only RCT of GH therapy in GHD, a short-term study for 1 year, children in the GH-treated group grew 2.7 cm/year faster than untreated children, with a statistically significantly higher height SDS: −2.3±0.45 versus −2.8±0.45.24 In uncontrolled larger studies treated final height SDS was −0.4 to −0.925 while near final height was 0.2–0.8 SDS below target height.26 There was no difference in final height between patients with isolated GHD and spontaneous puberty and those with multiple pituitary hormone deficiencies and induced puberty.25 There also appears to be no benefit in increasing the GH dose during puberty to mimic physiological increases,27 or attempting to inhibit epiphyseal fusion with co-administration of gonadotropin-releasing hormone analogues.28

Turner syndrome

It is the commonest gonadal dysgenesis in females, and is due to abnormalities within the X-chromosome. Just over 50% have a missing complete X-chromosome (45,X); the remainder have other abnormalities involving the X-chromosome including deletions of the short and long arms, duplications (isochromosomes) and ring chromosomes, or are mosaics. These abnormalities all act via SHOX gene haploinsufficiency. TS occurs in approximately one in 2000–2500 live female births, although it is estimated that up to 98% of affected fetuses spontaneously miscarry.

The short stature in TS is multifactorial, due to a combination of: intrauterine growth retardation, poor growth in childhood (characteristically from 4 to 5 years onwards), an absent pubertal growth spurt and a mild skeletal dysplasia, for example, short neck, wide carrying angle, hand and nail abnormalities. As a result final height is reduced by 21 cm from mid-parental height, and consequently mean untreated final height is 136–147 cm, although approximately 10% of patients have heights within the normal range.20

Patients with TS do not have classical GHD but abnormal GH secretion and GH insensitivity are both described,29 as is primary gonadal failure (hypergonadotrophic hypogonadism) with deficient oestrogen secretion.

In addition to GH therapy, other hormone therapies (anabolic steroids, sex steroids) are also often given, and so comparison of studies is limited by variations in therapy; doses and timing of therapy, and also lack of a control group. Short-term controlled studies over 1 year with either an untreated control group30 or low dose oestrogen treated group31 have shown an increase in HV of 2.4–2.6 cm/year, with an increase in height SDS of +0.3. A Cochrane review indicated that GH therapy improved HV by 3 cm/year in the first year of GH therapy, and 2 cm/year in the second year.32 In the one large study of GH therapy in TS where a placebo control group was used over 6 years,33 the improvement in final height was 7.6 cm.

Although studies indicate that the earlier GH is started the better the height outcome, the mean age of commencing treatment in the UK is 8.5 years, although this is a reduction in the 10.4 years seen in previous studies.34

In addition to GH therapy sex hormones (anabolic steroids and oestrogen) also impact on growth. A recent UK national randomised trial has demonstrated the benefit of addition of the mild anabolic steroid oxandrolone from 9 years of age onwards, but not delaying oestrogen replacement therapy given at 14 years of age rather than 12 years.35

Chronic renal insufficiency

The growth failure in chronic renal insufficiency (CRI) is multifactorial, arising from poor nutritional intake and anorexia, acidosis, anaemia, along with rickets.36 GH resistance, along with abnormalities of the GH/insulin-like growth factor 1 (IGF1) axis are also noted.37 Impaired growth occurs when the glomerular filtration rate falls below 50% of expected, and is significantly impaired below 25%.38

Although the prevalence of CRI is not known, it is estimated in the UK that 29% of children with CRI and renal transplants, and 41% of children on dialysis have heights <2nd centile.39 Despite this it is estimated that less than 5% of children with CRI receive GH therapy.

GH is recommended for prepubertal children with CRI, providing nutritional status and metabolic abnormalities have been optimised, and steroid therapy has been reduced to a minimum. GH treatment should be stopped after a renal transplantation, and only re-established after 1 year if it has been ascertained that catch up growth has not occurred.20

From four randomised 12-month trials, only one of which was placebo-controlled, HV increased by between 3.2 and 4.2 cm/year,40,–,42 with height SDS increases of +0.5 to +0.8,40 43 and a Cochrane review indicated that GH for 12 months produced a 3.8 cm/year increase in HV compared to untreated patients.43 Estimates of improvement in final height range from +0.5 to +1.7 SDS.

Prader–Willi syndrome

This is a dysmorphic syndrome with clinical features including short stature, hypogonadism, obesity, abnormal body composition, hypotonia, hyperphagia and learning and behavioural problems. It is due to loss of paternally derived genes on 15q, due to deletion in 70%, but also maternal uniparental disomy, translocations and imprinting defects.44

The incidence varies in a number of studies, with birth incidence in the UK estimated at 1 in 20 000, and a population prevalence of ∼1 in 50 000.45 Untreated final height is ∼154 cm in males and 145–149 cm in females.

It is unclear whether these patients are truly GH deficient or not, although PWS does share some phenotypic features: short stature with poor growth, abnormal body composition (decreased muscle mass, increased fat mass) and hypothalamic dysfunction (hyperphagia, central hypogonadism).

In PWS the aim of therapy is both to improve body composition as well as promoting growth, and GH results in improvements in both height (approximately 1 SD in the first year), body composition and muscle strength/tone. A number of short-term studies using untreated controls have demonstrated increases in HV over 12 months of 5.1 cm,46 47 and in height SDS of +0.7 to +1.4.46,–,52 Adult height SDS in uncontrolled treated groups ranges from −0.3 to −1.0 SDS. (http://bsped.org.uk).17

Four short-term controlled trials have shown statistically significant reduction in body fat, and improvement in lean mass.46 47 53 54

There have, however, been several reports of sudden death in PWS patients treated with GH, especially if patients are severely obese, and in these patients sleep studies should be performed before GH therapy is commenced.55

Small for gestational age

Although a variety of definitions exist, SGA is usually defined as a birth weight and/or length more than −2.5 SDS below the mean. These patients are a heterogeneous group, including normal children, and those growth retarded by maternal, placental and fetal factors.56 Approximately 80% of children born SGA show catch-up growth in the 6 months of life,57 which is completed by 2 years of age in most,58 although this may be delayed until up to 4 years of age in patients born both SGA and premature.59 Overall, approximately 10% of patients born SGA remain short (height <−2 SDS).60 In the UK and Europe GH is licensed for children born SGA (birth weight and/or length below –2SD (second centile)), who fail to show catch-up growth (HV SDS <0 during the last year) by 4 years of age or later, and who are short both compared to their peers (height <−2.5 SD) and parents (parental adjusted height <−1 SD). Although not part of the SGA licence, an International SGA Advisory Board recommended that GH treatment should be considered in children aged more than 4 years of age who show no catch up at a height −2 SD or less.56

In controlled trials treatment with GH for 1 year has produced an increase in HV of 1.3–2.9 cm/year and HV SDS +1.0 to+1.2.61,62 Other uncontrolled studies have shown a doubling of HV in the first year of GH treatment, with an increase in height SDS of +2.0 in the 3-year study.63 Only one controlled trial has been performed (in late treated children: mean age 12.9 years) followed to final height, with a height increment of 4 cm (0.6 SDS),64 although other uncontrolled studies with younger patients treated using higher doses of GH have shown height increments of up to 2.1 SDS.65

SHOX gene deficiency

This gene is located on the distal ends of both the X and Y-chromosomes, and plays an important role in long bone growth. As normal growth requires two functional copies, haploinsufficiency results in short stature. SHOX deficiency is demonstrated in a number of different conditions including 60–100% of patients with the skeletal dysplasia Leri-Weill dyschondrosteosis (see below),66 TS (see above), but also in 2.5–12.5% of patients with ISS (see below) who show no dysmorphic features.66 67 In the one controlled trial, GH-treated children had an increased HV of 3.5 cm/year in the first year and over 2 years gained 5.9 cm (0.9 SDS) in height.68 Preliminary data on final height in small numbers of patients from this study indicates a similar height SDS gain to TS (+1.1±(−0.7) vs 1.2±(−0.8)).69

Noonan syndrome

The incidence of Noonan syndrome (NS) is quoted to be from 1 in 1000 to 1 in 2500.70 Height of children with NS approximates the third centile in childhood, after which it usually falls further due to a delayed puberty and a decreased pubertal growth spurt. Overall mean adult height is approximately on the second centile (−2 SDS).71 Although retrospective studies (often using low doses of GH in children treated at an older age) demonstrate short-term improvement in growth with little improvement in final height,72 73 prospective studies have suggested a height SDS increase of +1.3 SDS74 and overall some studies have also indicated improved growth in NS patients with PTPN11 mutations.74

Idiopathic short stature

This is a heterogeneous group of children with a height 2 SDs or more below the mean for age (∼2nd centile), and in whom no endocrine, metabolic or other diagnosis can be made. A proportion, however, have been shown to have SHOX deficiency (see above). To be eligible for GH therapy in the USA, where there is a product license, patients must have a height below −2.25 SD (∼1st percentile). A recent Cochrane review75 showed that GH therapy can increase short-term height (up to 0.7 SD in 1 year) and improve (near) final height (7.5 cm and 3.7 cm taller than untreated and placebo treated controls in two studies). Increases in height are, however, such that treated individuals remain relatively short when compared with normal stature peers.

Skeletal dysplasia (achondroplasia)

There are probably several hundred different forms of skeletal dysplasia, and although final height varies between the different disorders it is often of the order of 110–130 cm. The most prevalent form is probably achondroplasia, which has an incidence of 1 in 25 000,76 and is almost always due to a consistent mutation in the FGFR3 gene.77 Studies indicate height increments of +1 to +1.5 SD in height over 5–6 years,78 79 although final height SDS is not significantly different from pretreatment height SDS.80 The only skeletal dysplasia which has a license for GH is Leri-Weill dyschondreosis (see SHOX deficiency above), and the remainder do not appear to be amenable to GH therapy.

Transition of GH treated patients

International consensus guidelines have been produced for GHD patients.81

These recommend that all patients should have re-evaluation of pituitary function at the completion of growth, with the extent of re-evaluation depending on the likelihood of profound GHD on retesting. Although discussion of the adult GH licence is beyond the scope of this article, patients profoundly GHD on retesting (<3 μg/l) should be offered GH therapy until at least 25 years of age to promote bone accretion and quality of life.

While NICE have previously recommended that all patients previously treated with GH should have long-term surveillance, in practice this often does not occur. Previous audits of GH treated patients in the UK14 still show quite wide variation in practice even in patients with GHD, including transfer, with 50% of units not transferring GHD patients normal on retesting. Although women with TS were being transferred to adult care, only a minority were being transferred to the recommended multidisciplinary adult TS clinics.82