Chapter 2: Normal Growth and Growth Disorders

Growth is a dynamic process influenced by many intrinsic and extrinsic factors that interplay to determine not only ultimate attained height but also the tempo and timing of increase in height. Research continues to unravel hormonal and genetic complexities that account for variations in “normal” growth, and etiologies for disordered growth. Careful tracking of childhood growth is a sensitive indicator of health and well-being, and therefore an essential component of sound pediatric care. Detection of unexplained acceleration or deceleration in growth rate or tracking along a disparate percentile considering family height genetics should prompt investigation. Endocrine disorders comprise an important, but only partial, differential diagnosis of abnormal growth. This chapter discusses essential components of normal growth, the detection and evaluation of worrisome growth, and diagnosis and treatment of its multiple etiologies. The ultimate goal is to provide for those caring for children a conceptual framework for the assessment of and diagnostic approach to the child with abnormal growth.

Phases of Normal Growth

The rate of linear growth and the physiologic components regulating it vary with age. Conceptually, it is helpful to define growth as occurring in four discrete but congruent phases—prenatal, infancy, childhood, and adolescence. According to the infancy-childhood-puberty growth model, phases are successive and partly superimposed. The infancy phase has been assumed to begin at mid-gestation and to tail off at approximately 2 to 3 years of age. It is regarded as being predominantly nutrition dependent. The childhood growth phase overlaps the infancy phase, begins sometime between 6 and 12 months of age, and is typically abrupt in onset. Its starting point probably defines the age at which intrauterine factors recede and growth hormone (GH) begins toexert significant influence (Figure 2-1).¹ It has been shown that the timing of infancy-childhood transition is significantly associated with final adult height, and a delayed transition is the most common predictor of later idiopathic short stature (ISS).² The transition from the childhood to puberty phase is characterized by an abrupt increase in sex hormones with a concomitant increase in GH secretion, and the contribution of the changing growth rates during puberty is reflected by the sigmoid pattern of height gain. The considerable individual variation in timing of puberty between early- and late-maturing children is depicted by brown shaded areas in Figure 2-2.

Early detection of deviation from normal growth velocity is the key to prompt evaluation and diagnosis of a child with a growth abnormality. Intrauterine growth occurs at a rate of approximately 1.2 to 1.5 cm per week, peaking at mid-gestation (18 weeks) at 2.5 cm per week and slowing to 0.5 cm per week just before birth. Intrinsic and extrinsic factors both contribute to prenatal growth and ultimate birth weight and length. Future growth patterns and genetic height tendencies, however, are not necessarily reflected at this stage. Hormonal regulation is largely by insulin and insulin-like growth factors-1 (IGF-1)and -2 (IGF-2) (see Figure 2-1). Discussed in more detail in the following discussion, maternal and uterine factors affecting fetal nutrition, insulin availability, and insulin sensitivity also have profound effects on intrauterine growth. Paternal size has relatively little influence on fetal weight gain, but paternal and maternal heights contribute roughly equally to fetal length gain. By comparison, growth hormone (GH) and thyroid hormone have only modest effects on in utero somatic growth (see the following discussion).

After birth, the rate of growth is simultaneously the most rapid and rapidly slowing of a child’s growth experience. This rapid slowing in growth rate coincides with diminishing postbirth sex steroid production and influence of nutrition-responsive IGF-1 effects to primary dependency on GH (see Figure 2-1). The average normal growth rates include 2.5 cm/month (0-6 months), 1.25 cm/month (7-12 months), 10 cm/year (12-24 months), 8 cm/year (24-36 months), and 7 cm/year (36-48 months). Growth rate continues to steadily decelerate through 4 years of age when the child normally transitions to the childhood phase characterized by an average growth rate of 5 to 6 cm per year (range 4.5-7.0 cm). Crossing of percentiles on the growth curve is not uncommon during the first 2 to 3 years of life. Genetic “rechanneling” (downward crossing of percentiles of a large baby [eg, affected by gestational diabetes] who is born to genetically short parents, or upward crossing of percentiles by a small baby [eg, exposed to an adverse prenatal environment] who is born to tall parents) is typically accomplished by 9 to 12 months of age. During 12 to 28 months, a tendency toward delayed growth is often manifest by growth rate deceleration that is more profound than average, resulting in down-channeling on the length-versus-age growth chart. After 36 months, normal growth follows a consistent percentile channel during the elementary school years. A prepubertal child whose growth velocity is less than 5 cm/year, whose height is less than 1st percentile, whose growth trajectory differs significantly from the family tendency, or who displays abnormal body proportions should be monitored closely and potentially evaluated.³

Depending on the age of puberty onset, growth rate may accelerate (early puberty) or decelerate (delayed puberty) compared to peers. Adolescence is marked by a return of rapid growth with peak sustained growth rates of up to 13 cm/year in boys and 11 cm/year in girls. The timing and tempo of puberty and rate of skeletal maturation differ between boys and girls; that is, the pubertal growth spurt occurs approximately 1.5 to 2 years earlier in girls than in boys, but boys grow on average 13 cm taller prior to fusion of the epiphyseal growth plates. The greater average final height in boys is in part due to this extended time in the prepubertal phase allowing for a longer period of growth prior to adolescence and partially due to the greater rate at which boys accrue height during their pubertal growth spurt.

The average onset of puberty differs not only between boys and girls but also among individuals of the same sex. The timing of sexual maturation plays a key role when assessing normalcy of growth rates (ie, determining expected growth rates for degree of skeletal maturation) and formulating height prognoses, especially when comparing an individual to population averages. The relatively broad timing of the normal onset of puberty (boys ∼9.5-14 years and girls ∼8-13 years) can cause a “late bloomer” to move downward in growth percentiles relative to their peers who are entering into the accelerated growth phase of puberty. Most of the disparity in normal adult height is accounted for by differences in growth rate (and, therefore, height achieved) prior to puberty; thus, in most cases, height at the time of puberty onset is a better predictor of a child’s final height than is age at onset of puberty.

Measurement

The most important tool in analysis of a child’s growth is the plotting of accurate measurements on an appropriate growth chart. Although a seemingly simple process, correct positioning of a child for accurate length or height during assessment and accurate data recording are critical and should be done by a trained individual. Recumbent length refers to stature taken while lying down. Recumbent length, used for infants and children younger than 3 years, is measured in supine position with full leg extension and with the head in the “Frankfurt plane” (Figure 2-3). This is best done using a firm box with a head plate and a movable footboard. If the child is older than 2 years and is physically able to stand on his/her own, a wall-mounted stadiometer is used for height measurement. The child, standing as erect as possible, should have four contact points: the back of his/her head, shoulder blades, buttocks, and heels, on the vertical plane of the stadiometer (Figure 2-4). The head should be aligned with a horizontal line from the ear canal to the lower border of the orbit of the eye, parallel to the floor and perpendicular to backboard. Importantly, there is discontinuity in the growth curves at age 2, which reflects the difference in recumbent length and standing height. Both length and height should be measured three times with no more than 0.3 cm variation and the mean recorded. Serial measurements taken to assess growth velocity should be obtained by the same individual to eliminate variations between examiners or equipment. Recording of data over 9 to 12 months is preferable to minimize the effects of measurement error and seasonal variations. When 4- to 6-month intervals are used, extrapolation to an estimated “annual growth velocity” is necessary for comparison with growth velocity charts.

Body proportions, like growth velocity, also change with age; normal large-headed and short-limbed newborns gradually transition by late childhood to “adult” proportions, with arm spans roughly equivalent to height and upper-to-lower (U/L) body segment ratios approximating one. Occipital-frontal head circumference, U/L body segment ratio (details about determination of U/L ratio in Table 2-1A and B), and arm span are useful in the assessment of short stature, tall stature, and markedly delayed or disproportionate growth. Detection of an U/L ratio above expected is characteristic of short stature owing to some genetic conditions (eg, Turner syndrome [TS] and certain chondrodysplasias), whereas a ratio below expected is observed in both some short-statured (eg, after spinal irradiation and spondyloepiphyseal dysplasia) and tall-statured (eg, Marfan syndrome) children.

Growth Charts

Measurement of the child (to the tenth of a centimeter) is plotted accurately against the percentiles of the population represented in the given chart. To construct a useful summation of a child’s growth, at least annual measurement and plotting of height-for-age is recommended. Calculation and plotting of growth velocity for age on reference charts (www.cdc.gov/growthcharts) allows for comparison of short-term growth with normal standards.

Growth charts should always be selected based on gender, whether height or length will be plotted, and, if relevant, any underlying syndrome. Both standing height- and length-for-age charts include data for children 2 to 3 years. As noted previously, because of normal substantial differences between an individual’s length and height, it is critical that height and length data only be plotted on height- and length-for-age growth charts, respectively. Otherwise, a misleading apparent discontinuity in the child’s percentiles may prompt unnecessary concern and evaluation.

A variety of genetic conditions affect growth, and there are specialized charts that may be considered for use with children affected by these conditions. These specialized growth charts may have some limitations. First, they are developed from relatively small homogeneous samples and, second, syndrome-specific growth charts may no longer reflect the current population. For example, for Down syndrome, it is now recommended that pediatricians use standard growth charts of the National Center for Health Statistics (NCHS) or World Health Organization (WHO) until new standards are developed.⁴ Alternatively, it is possible to plot the growth patterns of these children on both the specialized charts and the Centers for Disease Control and Prevention (CDC) growth charts. This will allow comparisons of growth to the general population of children and to the references for children identified with a given condition. The following link may be useful for several syndromes: http://www.dhs.wisconsin.gov.

In 2000, the NCHS (now part of the CDC) published charts based on cross-sectional data from physical examinations across the entire country (via the National Health and Nutrition Examination Survey [NHANES]) better reflecting cultural and racial diversity (www.cdc.gov/growthcharts). The percentiles plotted in these charts range from the 3rd to the 97th. For children who are growing outside of that range, the degree of short stature can be described by computing standard deviation scores (SDS) from the NCHS data: height SDS for age = (child’s height − mean height for age and sex)/SD for height for age and sex.

In 2006, the World Health Organization (WHO) released a new international growth standard for children aged 0 to 59 months. The CDC and WHO growth charts differ in their overall conceptual approach to describing growth. The WHO charts are longitudinal growth standards that describe how healthy children should grow under optimal environmental and health conditions. The curves were created based on data from selected communities worldwide, which were chosen according to specific inclusion and exclusion criteria.⁵ In contrast, the 2000 CDC growth charts are cross-sectional, and describe how certain children grew in a particular place and time (the United States during a span of ∼30 years [1963-1994]). Furthermore, whereas the CDC/NCHS growth charts for infants are derived from a sample of predominantly formula-fed infants, the WHO developed child growth standards for infants and young children based on the growth of breastfed infant as the biological norm. In 2010, the CDC recommended that the WHO growth charts be used in the United States for infants and children younger than 24 months, regardless of the method of feeding (Figure 2-5). When the WHO growth charts for formula-fed infants are used, normal and appropriate weight gain may appear to be too slow in the first 3 months of life and too rapid thereafter.

Although cross-sectional data are represented on standard height-for-age charts and are useful in assessing growth during the infant and childhood phases, they neither take into account nor depict normal variations in tempo of growth and maturation that influence timing of pubertal growth acceleration and growth termination. Height-for-age (Figures 2-6 and 2-7) and growth velocity charts based on longitudinal data that also incorporate sexual development, such as those produced by Tanner and colleagues, are more useful when comparing late- or early-maturing adolescents to population standards. These charts help the examiner take into account the relatively broad timing of pubertal onset and the acceleration of growth associated with the sexual development in an individual child.

Skeletal Maturation

Assessment of skeletal maturation provides information about the contribution of slowed or accelerated tempo of growth to a child’s growth pattern, and is, therefore, another key component in the evaluation of an abnormally growing child. The multiple physiologic factors involved in growth plate elongation and maturation are still being unraveled, but it is currently well established that calcitropic hormones, GH, IGF-1, thyroxine (T4), sex hormones, glucocorticoids, and C-natriuretic peptide play key roles in this process.

The rate of longitudinal bone growth is a product of the rate of differentiation and production of new growth plate chondrocytes times the average size of the terminal hypertrophic chondrocytes that form the scaffold for bone elongation. Evidence supports dual, IGF-1-independent and IGF-1-dependent, roles for GH in promoting longitudinal bone growth. GH enhances chondrocyte generation and proliferation independently of IGF-1, and IGF-1 appears to promote chondrocyte hypertrophy.⁶

While both androgens and estrogen are associated with acceleration in growth plate elongation and maturation, it is now known that estrogen has the predominating effect on this process. Estrogen (produced from androgens via the aromatase enzyme) controls the timing of the growth spurt and the conversion of proliferating into hypertrophic, but senescent, chondrocytes in the growth plate (and, hence, progress toward closure and bone mass accretion). This phenomenon is an important pubertal process as 90% of skeletal mass is accrued by 18 years of age. Both estrogen receptors, ER-α and ER-β, are expressed in the growth plate in boys and girls throughout pubertal development. Estrogen has biphasic actions on longitudinal growth in both girls and boys: Very low levels of estrogen stimulate bone growth directly at the growth plate and indirectly through increased GH secretion and IGF-1 production and action. Conversely, higher levels of estrogen inhibit GH action at the level of receptor expression and signaling (specifically inhibiting Janus kinase–mediated transcription via induction of SOCS2 [suppressor of cytokine signaling-2]) and stimulate epiphyseal fusion. Androgens regulate both proliferation and differentiation of cultured epiphyseal chondrocytes and, thus, also have direct effects on growth plate cartilage and longitudinal bone growth. In addition, though androgens indirectly increase GH secretion and pulsatility due to aromatization to estrogen, a direct effect of androgens on epiphyseal growth and maturation is supported by the observed growth rate acceleration observed with oxandrolone, a nonaromatizable androgen that does not affect GH secretion.

Skeletal maturation is assessed by examination of a “bone age” (BA) film, which is a radiograph of the left hand and wrist. Because the ossification of trabecular bone occurs in a predictable pattern, the evaluator can quantitatively assess the bone maturation of multiple ossification centers and compare it to standard male or female radiographs of different ages. The BA can then be compared to the patient’s chronological age as an indicator of growth tempo, remaining growth potential, and a judgment of statistically abnormal delay or advancement (with use of normative ± SD reference data). Assessment of BA is particularly useful in children with the following conditions: short stature, tall stature, early or late puberty, and congenital adrenal hyperplasia. In children younger than 2 years, given the limited number and size of visible ossification centers in the hand, a hemiskeleton BA can be obtained by obtaining radiographs of the child’s entire left upper and lower extremities. The number of ossification centers is determined and again compared to age-appropriate standards to determine the bone age.

Although indispensable in the evaluation of growth, the use of a BA in practice also has its perils. The single film of the left hand is assumed to be representative of ossification centers in general and, thus, avoids radiation exposure required for radiographs of the entire skeleton. However, because the hand does not contribute to height, its accuracy in predicting growth potential can be limited. Most commonly, the radiograph is compared with the published standards of Greulich and Pyle (G-P),⁷ a set of normal age-related standards that were collected from a small cohort of American white children between 1931 and 1942. Consequently, use of these dated standards has dwindling accuracy not only for the assessment of (faster growing and maturing) children in the 21st century but also of children from other ethnicities or children with intrinsic disease processes such as skeletal dysplasias. Nonetheless, the G-P atlas is still the most widely used skeletal development standard. An alternative Tanner-Whitehouse method for assessing BA from radiographs of the left hand involves a scoring system for developmentally identified stages of each of 20 individual bones, a technique that has been adapted for computerized assessment.⁸ Further, assessment of BA can have substantial variability of intraobserver interpretation. It is, thus, optimal to have a consistent reviewer, especially in the context of serial BAs in the same patient.

Finally, radiographs of the hand can be helpful in detection of many skeletal disorders such as Madelung deformity characterized by arrest of epiphyseal growth of the medial and anterior portions of the distal radius leading to shortening of the radius and relative overgrowth of the ulna (Leri-Weill dyschondrosteosis and TS) and short fourth metacarpal (pseudohypoparathyroidism and TS).

Prediction of Adult Height and Parental Target Height

The extent of skeletal maturation can be used to predict ultimate height potential. Several methods have been developed for the prediction of adult height (PAH) (Table 2-2). All are based on the premise that delay in BA relative to the chronological age is proportionate to growth potential remaining. For each method, as BA advances, the PAH becomes more accurate. The most commonly used one was developed by Bayley and Pinneau and is based on G-P Radiographic Atlas of Skeletal Development. This calculation takes into account the BA, chronological age, and current height, and a semiquantitative allowance for chronological age. Importantly, all methods of calculating PAH are based on data from normal children, and none has been documented to be accurate in children with growth abnormalities. There are several alternative methods for PAH such as those described by Tanner-Whitehouse (TW) and the Roche-Wainer-Thissen (RWT) (Table 2-3). These methods have a common shortcoming as they are based on manual BA determination. Automated methods for BA determination have been recently developed.⁹ In addition, several models have been developed for specific situations such as constitutional delay of growth and puberty (CDGP), short normal children, tall stature, and congenital adrenal hyperplasia.

Assessment of the PAH is frequently used in conjunction with the mid-parental target height (MPH), which takes into account familial genetic factors in growth and height potential. The MPH calculation is the average of the mother’s and father’s height adjusted by the average height difference of 13 cm between the sexes. Thus, if the MPH of a female patient is calculated, 13 cm is subtracted from the father’s height before it is averaged with the mother’s. For the evaluation of a male patient, 13 cm is added to the mother’s height before averaging the parents’ heights. In parts of the world where average heights continue to increase, a generation “correction factor” is also included. Like the PAH, this is a broad estimation (ie, 2 SD range for this calculation is ±10 cm), but it can be useful in comparing a child’s growth to that of his/her parents and siblings. The MPH is also useful for identifying whether a child’s current and predicted growth trajectory differs from that expected; that is, a significant disparity in the child’s growth percentile (ie, ±10 cm) from the MPH percentile should prompt further investigation (most often revealing constitutional early or delayed growth and development) even if the child plots within the normal percentile range. Additionally, when there is a relatively large disparity between the parents’ heights, the calculated MPH becomes less informative. While a child’s growth can be disproportionately influenced by the pattern of one parent, statistical reassessment of population averages shows that the final height of offspring tends to regress toward the mean. If this is not taken into account, the shorter calculated MPH may be used to inappropriately explain the short stature of a child.

Endocrine Regulation of Growth

Regulation of growth is provided by a complex interplay of stimulatory and inhibitory hormones influenced by afferent messages from the central nervous system (CNS) and periphery. The function of these systems reflects genetic programming, the nutritional and developmental state of the body, and the presence of interfering processes. Thus, evaluation for “endocrine” causes of abnormal growth includes, in addition to assessment of the GH/IGF-1, thyroid, and, as appropriate, gonadotropin/sex steroid axes, an exploration for nutrition deficiencies, psychosocial stress, growth-interfering medications, and occult organ dysfunction (eg, renal or gastrointestinal disease). Regulation of somatic growth by the endocrine system varies significantly during a child’s growing lifetime.

Regulation of fetal growth

Fetal growth regulation involves complex interactions between the mother, placenta, and fetus. The placenta is the site of bidirectional exchanges receiving nutrients and oxygen from the mother and transferring fetal waste from the fetus to the mother. Therefore, adequate placental growth and function are essential for normal fetal growth. However, the placenta is also a complex endocrine organ sending signals to the mother that influence maternal metabolism and, therefore, fetal growth. The placenta signals also to the fetus by secreting steroid hormones and human placental lactogen (hPL). While estrogen and progesterone seem to have no role in fetal growth, hPL may promote fetal growth by stimulating other fetal hormones such as insulin or IGF-1. The placenta is also a barrier, protecting the fetus from negative influences coming from the maternal compartment. The maternal cortisol level is 5 to 10 times higher than the fetal cortisol concentration, a difference made possible by a placental enzyme, 11β-hydroxysteroid dehydrogenase-2 (11βHSD-2) that converts cortisol into its inactive 11-keto-metabolite, cortisone. By doing so, this enzyme protects the fetus from deleterious effects of high cortisol concentrations. Several studies have shown that a reduced 11βHSD enzyme activity is associated with reduced fetal growth.

In the fetal compartment, the key regulators of fetal growth—insulin, IGF-1, and IGF-2—are produced and secreted by the fetus early during fetal development. Maternal insulin does not cross the placental barrier and, thus, the fetus depends on its own production for the control of glucose metabolism. Insulin, produced by β-cells in the fetal pancreas beginning at the end of the first trimester, is an important regulator of fetal growth and development. The β-cell is responsive to glucose stimulation during the second half of gestation and the amplitude of insulin response increases during gestation. There is a correlation between increase in insulin production and increased fetal growth, and thus any variation in glucose availability to the fetus will influence fetal insulin secretion and fetal growth. Deficiency in insulin production (eg, severe neonatal diabetes mellitus [DM] and pancreatic agenesis) or action (eg, leprechaunism) results in reduced fetal growth. In fetuses and neonates born small for gestational age (SGA), cord blood insulin concentrations are significantly lower than appropriate for gestational age infants. Conversely, in response to maternal hyperglycemia, fetal insulin production increases, leading to increased birth weight and macrosomia observed in infants of poorly controlled diabetic mothers. This increased birth weight is largely due to increased body fat acquisition and the effect on birth length is modest.

The IGF system also plays a central role in the regulation of fetal growth and normal local production, and action of IGFs appears to be indispensable for normal growth in utero. IGF-1 and IGF-2 are expressed in fetal tissue from zygote formation and implantation until just before birth. During fetal life, the IGF-1 concentration correlates positively with gestational age. Normal fetal production of IGF-1 is regulated primarily by nutrition and, importantly, is largely independent of GH secretion. As a result, neonates with GH deficiency (GHD) or insensitivity, as noted previously, have normal to low-normal length at birth, followed by marked postnatal growth failure (Figure 2-8).¹⁰

In contrast, infants with primary over- or underproduction of IGF-1 demonstrate marked effects on in utero growth, exemplified by disruption of normal growth of patients displaying IGF-1 resistance owing to mutations within the IGF-1 gene or its receptor (IGF1R). These patients show marked intrauterine growth restriction (IUGR), are born SGA, and fail to catch up in length as infants.¹¹ Further, IGF-1 levels in fetal and cord blood correlate with fetal size, and, in spite of higher GH levels, they are reduced in SGA infants (Figure 2-9).¹² IGF-1 continues to be critical for normal linear growth in postnatal life. While IGF-1–binding proteins (IGFBPs) are identifiable in the fetus and newborn, levels of IGFBP-3 and acid-labile subunit (ALS,) the primary carriers of IGF-1 in the circulation after infancy, are low.

IGF-2 expression is more extensive in fetal tissues than is IGF-1 from mid-to-late gestation in rodents and humans. An important role of IGF-2 in fetal growth is illustrated by the growth anomalies observed in Russell-Silver syndrome (RSS) and Beckwith-Wiedemann syndrome (BWS). Phenotypically, RSS is characterized by severe intrauterine and postnatal growth retardation, whereas BWS is characterized by fetal overgrowth in late pregnancy and later during early childhood. The molecular basis is complex and involves abnormalities in the IGF-2 region on chromosome 11p15. It is believed that the excessive growth observed in BWS is due to overexpression of IGF-2 and excessive production during fetal life, and the reverse appears to be true in 30% to 65% of cases of RSS. Unlike IGF-1, in normal conditions, IGF-2 does not appear to play a critical role in postnatal somatic growth.

In contrast, GH seems to be unimportant for fetal growth. GH is secreted during fetal life beginning at mid-gestation and its concentration at birth is high compared to what is observed later in infancy. However, infants with severe neonatal hypopituitarism, despite profound GHD, have a normal or low-normal length at birth but have a decline in their growth velocity during the early months of life. Similarly, neonates with the GH insensitivity (GHI) syndrome (Laron dwarfism) have only a slightly reduced birth length but show a rapid decline in their growth velocity during the early months of postnatal life. These examples illustrate that GH has a limited role in fetal growth (and IGF-1 production) but begins to influence linear growth during early infancy. Thyroid hormone, another hormone critical for normal postnatal growth, influences bone mineralization and maturation in utero but does not substantially influence fetal growth. Thus, infants born with severe congenital hypothyroidism demonstrate normal length but have delayed bone maturation.

Regulation of postnatal growth

During the first few months of infancy, nutritional adequacy (and its effects on insulin and IGF-1 production) continues to exert strong influence on linear growth, so that infantile GHD may not always be manifest by growth failure at this time. Thyroid hormone deficiency, while of relatively little importance to growth of the fetus, can cause profound postnatal growth failure and virtual arrest of skeletal maturation. In addition to having direct effects on epiphyseal cartilage, thyroid hormones appear necessary for normal GH secretion. Patients with hypothyroidism have decreased spontaneous GH secretion and blunted responses to GH stimulation tests which normalize with restoration of euthyroidism.

Childhood growth is primarily regulated by GH, normal secretion and action of which is dependent on normal thyroid hormone function and adequate nutrition. Gradually declining growth rates during childhood may be matched by declining laboratory evidence of GH secretion, particularly when pubertal onset is delayed. This phenomenon can create difficulty in distinguishing physiologic growth deceleration from true GHD during the immediate prepubertal years. Normal growth acceleration and completion during adolescence is dependent on actions of estrogen and androgen, their effects on GH secretion, and normal thyroid hormone and nutritional status. Gonadal steroids produce a growth spurt partially by enhancing GH secretion (via aromatization of testosterone to estrogen in the CNS), and also by stimulating IGF-1 production and chondrocyte proliferation in the growth plate directly. Nevertheless, without GH, an adolescent will not have a normal pubertal growth spurt.

In addition to regulating growth, GH has important metabolic effects. These include (1) stimulation of bone remodeling (by stimulating both osteoclast and osteoblast activity) with ultimate net increase in bone mass;(2) stimulation of lipolysis and fat utilization for energy expenditure; (3) growth and preservation of lean tissue mass and muscle function; and (4) facilitation of normal lipid metabolism (see Figure 2-23). Sex-specific dimorphism in these metabolic effects, reflecting underlying sex steroid–mediated differences in GH sensitivity and secretion, becomes apparent during adolescence leading to females having higher GH levels than do males. In animal models, sexually dimorphic GH release from the pituitary is characterized by larger and more regular pulses in males and smaller, more frequent, and more irregular pulses in females. There are also sex differences in downstream signaling pathways in target tissues, such as sex-specific gene expression in the liver.¹³ Beyond linear growth, normal production of GH is needed for the adolescent to accomplish normal body composition maturation. Adults continue to secrete GH pulses, but the amplitude decreases with age, returning to a pattern more characteristic of the prepubertal child in early adulthood followed by a gradual, but steady, decline with aging.

The GH/IGF-1 axis

Growth hormone synthesis and secretion.

While optimal growth and development occur only with the normal production and modulation of multiple hormones, the predominant common pathway for the endocrine regulation of childhood somatic growth is the GH/IGF-1 axis. GH synthesis and secretion occurs in/from the anterior pituitary (adenohypophysis), which originates from the floor of the primitive pharyngeal epithelium known as Rathke pouch. Infolding of this tissue occurs around days 15 to 20 of gestation and subsequently migrates caudally to meet a neural ectodermal extension from the floor of the midbrain that will develop into the posterior pituitary. These two distinct embryologic tissues combine to form the pituitary gland, although the functions of the anterior and posterior portions remain largely distinct.

GH is produced in the somatotrophs of the anterior pituitary as a single-chain 191-amino acid, 22-kDa protein. GH release and synthesis reflect a balance between the effects of three hypothalamic releasing hormones: GH-releasing hormone (GHRH; stimulatory), ghrelin (stimulatory), and somatostatin (also known as somatotropin release inhibitory factor [SRIF]; inhibitory). GHRH is produced in neurons that have their bodies in the ventral arcuate nucleus of the hypothalamus and axons projecting into the median eminence, allowing the GHRH to be secreted into the portal system of the pituitary stalk. GHRH, through interaction with a G-protein receptor, stimulates a cyclic adenosine monophosphate (cAMP)–protein kinase A activation pathway causing an influx of calcium ions into the somatotrophs and subsequent synthesis and release of GH. Counter-regulation of GH release is directed by somatostatin, produced in the periventricular nucleus of the hypothalamus and also released into the same portal system as GHRH. SRIF binds to a G-protein receptor that (1) acts on somatotroph K channels, preventing outflow of K ions needed to accommodate the influx of positively charged calcium ions, and (2) inhibits cAMP production. Through these actions, SRIF inhibits the release, but not synthesis, of GH from somatotrophs. Ghrelin is a synthetic hexapeptide secretagogue capable of binding to a unique receptor and stimulating GH secretion through effects on both the hypothalamus and pituitary. Ghrelin is secreted primarily from the stomach (and to lesser extent from small intestine and CNS) during fasting, implicating a role of GH release in the coordination of caloric expenditure and partitioning. Though ghrelin is capable of stimulating GH release, it does not appear to be involved physiologically in the immediate control of GH release at the level of the somatotroph. Centrally, ghrelin evokes a GH secretory pulse by increasing release of GHRH while also suppressing the inhibitory effect of SRIF at the level of the pituitary.¹⁴ The shifting balance between GHRH and SRIH determines the pulsatile release of GH, with SRIFinfluencing the timing and amplitude of GH pulses. GHRH release during SRIF secretion allows for GH synthesis, but not release from the somatotrophs. During times of decreased SRIF tone, stored GH is released in a pulse. Both GHRH and GH work in a negative feedback loop to stimulate SRIF release, and GH itself, in a short feedback loop, also inhibits GHRH.

Production and release of GH is influenced by multiple factors including neuropeptides, neurotransmitters, and hormones that can either increase or decrease GH release. GnRH, pituitary adenylyl cyclase activating polypeptide (PACAP), and thyrotropin-releasing hormone (TRH), together with peripheral peptide hormones such as leptin and adiponectin, stimulate GH release. Physical activity, sleep, fasting, and dopamine agonists also stimulate GH release. During stage IV sleep, SRIF activity is low while GHRH activity persists, resulting in nocturnal pulses of GH, the main contribution to daily GH secretion. These pulses increase in both amplitude and frequency during puberty when the effect of estrogen in both males and females augments the release of GH. IGF-1 inhibits GH release, and one growth-suppressing effect of persistent glucocorticoid excess is through inhibition of GHRH and augmentation of SRIF. This complex system of stimulatory and inhibitory feedback loops, which modulate GH production and secretion, is depicted in Figure 2-10.

Growth hormone action.

In serum, GH is either free or bound to the GH-binding protein (GHBP), which is derived from cleavage of the extracellular portion of the GH receptor (GHR). GHBP in human plasma binds GH with high specificity and affinity, but with relatively low capacity, as about 45% of circulating GH is bound. GHBP prolongs the half-life of GH, presumably by reducing its glomerular filtration, and modulates its binding to the GHR. Like GH, GHBP concentrations vary with age, estrogen levels, body mass index (BMI), and feeding states. At the target tissue cell membranes, steps in free GH action include (1) binding of GH to the membrane-associated class-1 cytokine tyrosine kinase GHR; (2) sequential dimerization of the GHR through binding to each of two specific sites on GH; (3) interaction of the GHR with Janus kinase 2 (JAK2); (4) tyrosine phosphorylation of both JAK2 and the GHR; (5) changes in cytoplasmic and nuclear protein phosphorylation and dephosphorylation; and (6) stimulation of target gene transcription. JAK2-dependent phosphorylation and activation have been demonstrated for many cytoplasmic signaling molecules that, after forming homodimers or heterodimers, translocate to the nucleus, bind DNA, and activate transcription of growth-promoting products, including IGF-1 (Figure 2-11).¹⁵ Particularly in the liver, activation of GHRs also initiates the signaling pathway for the production of IGFBP-3 and ALS. These components combine to form a 150-kDa complex that transports IGF-1 in the blood and prolongs its half-life. Precisely how these pathways mediate the various anabolic and metabolic actions of GH remains to be elucidated. Although most measurable IGF-1 comes from the liver, GHRs are expressed in multiple tissues (including growth plates) providing a pathway for GH to have effects on growth independent of the hepatic production of IGF-1. GHRs are found in epiphyses, prechondrocytes of the cartilage precursor cells, and in bone marrow. At these locations, autocrine and paracrine action of locally produced IGF-1 appears to be the prime stimulator of somatic growth and elongation of bone (based on genetic studies in mice). In other tissues, GH also has important body composition effects, resulting in the promotion of lean mass accretion. GH provides counter-regulation to insulin, leading to an increase in lipolysis and inhibition of lipogenesis in adipose tissue, as well as to an increase in amino acid transport and nitrogen retention in muscle.

Age-related issues in GH secretion.

The connection between the hypothalamus and the pituitary portal system is developing by 9 weeks of fetal life at which time somatotrophs can be identified in the anterior pituitary. By 12 weeks of life, the regulatory effects of GHRH and somatotropin release inhibitory factor (SRIF) are in place. GH is secreted readily from this point in gestation on but, as mentioned earlier, plays only a small role in fetal linear and weight growth. A number of recently identified genetic mutations can result in GHD, including HESX1 (thought to be linked with septo-optic dysplasia), SOX2, SOX3, GLI2, LHX-3 and LHX4 (pituitary hypoplasia), PROP1 or POU1F1 (combined pituitary hormone deficiency), OTX2, (critical role in retinal photoreceptor development), and GH1 (isolated GHD [IGHD]). (Table 2-4).¹⁶

Immediately after birth, GH levels are normally high through the first few days of life, with random values greater than 20 ng/mL being common. Because GH provides insulin counter-regulation (in conjunction with cortisol) during infancy and its release is normally stimulated by hypoglycemia, a GH level less than 20 ng/mL concurrent with hypoglycemia strongly suggests GHD. Because, as stated previously, GH is not the key regulator of linear growth velocity during the primarily nutrition-dependent phase of very early infancy, neonates with GHD may not show slowed linear growth until as late as 18 to 24 months of age.

During childhood, most GH secretion occurs in two to five discrete pulses during sleep. Because of this secretion pattern, GH levels usually range from low to undetectable on random blood samples. A wide range of spontaneous GH secretion has been observed in both normally growing and short children, so that testing this with frequent blood sampling has not proven diagnostically useful. Because the diagnosis of GHD usually does not connote a complete deficiency, affected children will often continue to grow, but at a subnormal rate that results in a gradual downward crossing of percentile lines and slow decline in height SD score.

During adolescence, a rise in estrogen leads to an increase in GH secretion and IGF-1 production (directly and indirectly). While this occurs in both sexes, higher levels of GH are typically observed in pubertal females, reflecting not only higher estrogen levels but also the appearance of sex-specific differences in GH-induced hepatic IGF-1 production and feedback. GH pulses during the adolescent phase increase in frequency to occur both during the day and night and have higher amplitudes. GH bioavailability is also increased because production of its binding protein, GHBP, remains stable.

The IGF system.

Two IGFs have been identified: IGF-1 and IGF-2. IGF-1 is a 70-amino acid peptide that is slightly alkaline and its sequence is encoded on the long arm of chromosome 12. IGF-2 has 67 amino acids, is slightly acidic, and encoded on the short arm of chromosome 11. Both peptides are composed of two amino chains connected by disulfide bonds. They share 45 amino acid positions and have approximately 50% homology to insulin. Depending on the tissue, there are multiple different messenger ribonucleic acid (mRNA) sequences as a result of variable splicing and regulation of transcription of the IGF-1 and -2 genes, adding another layer of sophistication to the effects of IGF expression and, thus, its growth effects.

The IGF1R (officially known as the type 1 IGF receptor) binds both IGF proteins with high affinity, and both IGF-1 and -2 activate the tyrosine kinase stimulatory intracellular cascade through the IGF1R. It is through this receptor that mitogenic and metabolic actions of IGF proteins are mediated. Like the GHR, the IGF1R is expressed in a multitude of tissues. The IGF1R is made up of two transmembrane α-subunits forming the IGF-binding sites and two intracellular β-subunits that make up the signal transduction component of the receptor. The IGF-2 receptor (IGF2R and officially known as the type 2 IGF receptor) only binds IGF-2 with high affinity and does not exert effects through the same tyrosine kinase activation pathway. The downstream effects of activating the IGF2R are predominantly antimitogenic rather than growth stimulating. This receptor pathway may be involved in regulating the inhibitory effects of retinoids and modulation of growth inhibition of the IGF system.

In addition to the IGF receptors, the IGFBPs also modulate IGF action. Six IGFBPs are produced and bind 70% to 80% of IGF-1 and -2. Their regulation by GH varies, with IGFBP-3 being closely stimulated by, and IGFBP-1 being essentially independent of, GH. The IGFBPs serve as carrier proteins for the IGF peptides, transport the IGFs to target cells, and modulate the interaction of the IGFs with their cell surface receptors. Highly conserved cysteine-rich regions of these carrier proteins are the sites of the disulfide bridges that play a key role in the structure and formation of the IGF binding site. When compared to IGF receptors, IGFBPs have a higher affinity for the IGF peptides. The bioavailability of free IGFs appears dependent on (and is further regulated by) the action of IGFBP proteases (Figure 2-12).¹⁷

Developmental changes in IGFs.

The IGF-1 levels vary substantially throughout life. At birth, levels are relatively low (∼30%-50% of adult levels) and are influenced by gestational age and weight (see Figure 2-9). The IGF system plays a central role in the regulation of fetal growth (see the previous discussion). While IGFBPs are identifiable in the fetus and newborn, levels of IGFBP-3 and ALS, the primary carriers of IGF-1 after infancy, are low. Thus, it appears that fetal IGF-1 production is linked more tightly to nutrient supply and to insulin secretion than to GH regulation. Levels subsequently rise during childhood, reaching peak levels right after the pubertal growth spurt (13-14 years in girls and 15-16 years in boys). Under the influence of gonadal steroids, both directly and indirectly via the increase in GH pulsations, IGF-1 levels will increase to two to three times the adult range (Figure 2-13). During the pubertal phase of growth, the levels of IGF-1 more closely correlate with Tanner stage and BA rather than with chronological age. IGF-2 levels show less age variation and, by 1 year of age, achieve adult levels with little subsequent change.

A normal growth rate is an important indicator of childhood health. Therefore, proper growth assessment is a key component of pediatric care, requiring both careful measurement of length or height at discrete points and tracking of growth velocity over time. Though accurate plotting of isolated linear growth measurements on growth charts will easily identify children who are abnormally short or tall in stature (ie, whose height falls outside ± ∼2 SD, < 3rd or > 97th percentile), early detection of a growth disorder frequently requires appreciation of both subtle changes in growth velocity or recognition that a growth pattern, while within the statistically normal range, is divergent from that expected for a particular child.

Evaluation of the child with worrisome growth begins by systematically asking and answering the following questions: (1) Is the growth rate normal or abnormally slow or fast? (2) What is the relationship between the linear growth rate and poor, normal, or excess weight gain? (3) What intrinsic, familial, or other genetic factors may be influencing this child’s growth? (4) What is the family history for pubertal onset and age of adult height attainment, and is there BA evidence for delayed or accelerated growth? Because the normal range within the population is fairly wide, skillful and sensitive analysis that uses the answers to these questions can help in determining a more narrowly defined “expected” normal range for a specific child. Appreciation of the influences of superimposed factors in a single child usually provides either rationale for reassurance and observation or reason to pursue a diagnostic workup and possible treatment.

• Is the growth rate normal or abnormally slow or fast? Growth rate is the change in length or height measured on separate occasions relative to the time elapsed. The normal range of growth velocities varies markedly with age, skeletal age, and pubertal stage. Growth rates estimated from measurements separated by months can be annualized to centimeters/year to allow comparison with normal growth rate charts for age (see Figure 2-2). To more precisely evaluate whether a growth rate is normal in children with delayed or accelerated growth patterns, determination of skeletal age and of “growth rate for BA” can also be helpful. Consistently diminished growth velocity becomes manifest as downward crossing of percentiles on the length- or height-for-age growth chart, a situation that demands evaluation even when length or height remains within the normal range. Thus, careful attention to growth rate, and not only length or height, facilitates early detection of a growth-slowing disorder in taller and shorter children alike.

• What is the relationship between the linear growth rate and poor, normal, or excess weight gain? Both poor nutrition and excess caloric intake can influence linear growth. Particularly, early in life it is diagnostically helpful to distinguish between “failure to thrive” (ie, weight deficit > length deficit) (Figure 2-14) and “failure to grow” (length deficit > weight deficit), as the former should prompt initial evaluation for nutritional problems or undiagnosed systemic illness (Figure 2-15), while the latter is more typical of growth-retarding endocrine disorders.³ On the other hand, overnutrition is a common cause of linear growth acceleration, with weight gain that precedes and exceeds in magnitude the increase in linear growth percentiles, usually accompanied by commensurate advancement in skeletal maturation and onset of puberty particularly in females (Figure 2-16). Endocrine causes of obesity such as hypothyroidism and cortisol excess need not be considered when a normal or accelerated growth rate accompanies weight gain.

• What intrinsic, familial, or other genetic factors may be influencing this child’s growth? Intrinsic and genetic influences on the childhood growth rate and height attainment encompass a wide array of factors whose distinctive underlying physiologies are slowly yet steadily being unraveled. The parental contribution to a child’s expected height percentile range is estimated using the MPH calculation described earlier. Important caveats include (1) diminishing accuracy of this calculation as parental height percentiles become more disparate and (2) the assumption that a parent’s height represents his/her own genetic “potential” (ie, was not affected by an early life growth-restricting influence). An extended family history is helpful in revealing sporadic growth disorders (eg, short-limb tendencies) that may not manifest in every generation. Determination of body proportions, in particular the upper-to-lower body segment ratio (corrected for age and BA, as this measurement normally varies throughout growth and development), may aid in the identification of some causes of short (eg, hypochondroplasia and TS) or tall (eg, Klinefelter and Marfan syndromes) stature. Awareness of birth length and weight to determine whether the child was born SGA is critical to assess whether restriction of intrauterine growth may be persisting into childhood. Careful physical examination may detect stigmata of known chromosomal disorders associated with short (eg, Turner, Noonan, Down, and Russell-Silver syndromes) or tall (eg, Marfan, Klinefelter, and Soto [early in childhood] syndromes) stature.

• What is the family history for pubertal onset and age of adult height attainment, and is there BA evidence for delayed or accelerated growth? The range of normal variations in “tempo” of childhood growth is not captured on most standard height-versus-age growth curves, but determination of this dimension is indispensable to evaluation of worrisome growth (variations in which determine the time required for an individual child to complete his/her growth) (see Figures 2-6 and 2-7). Differences in the time required to progress toward pubertal maturation and ultimate growth potential are influenced by biological variations in GH secretion and action, variations in suppression of the hypothalamic-pituitary-gonadal axis during infancy and its subsequent reawakening at pubertal onset, and other factors. A family history of early or late puberty or attainment of adult height is helpful in assessing the likelihood that a similar growth pattern in the child represents a normal variation. Helpful questions to elicit this history from parents include height at the beginning of high school, the age of maternal menarche, and continued paternal growth after high school. Objectively, the degree to which delay or acceleration in the growth process is contributing to a child’s position on the growth curve is estimated by examination of BA.

The astute clinician assembles this information and then asks: “to what extent do the historical and physical examination findings explain the child’s growth pattern and position on the growth curve?” Commonly, the process outlined previously reveals historical and familial short or tall stature traits that, when combined with tendencies toward delayed or accelerated tempo of growth and pubertal development, provide a reasonable explanation. Further evaluation of the differential diagnosis outlined as follows, however, should be prompted by (1) documented growth velocity less than 10th percentile or greater than 95th percentile for age (ie, any abnormally slow- or fast-growing child); (2) height-for-age less than 1st percentile (ie, any markedly short child); (3) height projection (current height percentile with correction for BA delay or advancement) that differs significantly from MPH; and (4) detection of abnormal body proportion.

Diagnosis and Treatment of Worrisome Growth

Growth disorders in childhood generally display one or more of the following characteristics: (1) short stature with normal tempo of growth (ie, growth rate and skeletal maturation within the normal range, familial/genetic short stature); (2) short stature with normal growth rate, but evidence for prior slowed tempo of growth (ie, “constitutional growth delay” with delayed skeletal maturation); (3) abnormal growth rate with or without short stature (attenuated growth caused by systemic disease or hormonal deficiency); or (4) growth acceleration. Proper interpretation and diagnosis of growth disorders frequently require appreciation of overlapping growth patterns in the same child. This is commonly observed in children with extreme “idiopathic (ie, nonpathologic) short stature” resulting from combined influences of familial short stature and constitutional growth delay. Sound therapeutic decisions regarding growth problems must incorporate a clear understanding of both the complexity of multiple factors simultaneously influencing growth in a child and the natural history of these growth patterns. In many situations, careful watchful waiting is a critical first “intervention” that allows distinction of normal variations in growth from disorders for which treatment is needed. A conceptual approach to the differential diagnosis of a child with worrisome growth is depicted in Figure 2-14. Box 2-1 lists the circumstances in which a patient with growth disorders should be referred to a pediatric endocrinologist.

Growth hormone therapy overview

For many years, treatment of growth-retarding disorders was confined to correction of underlying disease or hormonal disturbance. During the past 30 years, however, treatment of growth disorders has expanded beyond hormone-replacement therapy to include enhancement therapy to improve stature, body composition, and perhaps quality of life. One consequence of this expansion has been the evolution of complex ethical dilemmas in recombinant DNA-derived human GH (hGH) therapy regarding appropriate criteria for the initiation and discontinuation of therapy, responsible allocation of health care resources, and the growth of “cosmetic endocrinology.”³ Though hGH therapy has dominated as the growth-promoting therapy of choice, advances in the understanding of growth regulation and skeletal maturation by other factors (eg, estrogens and androgens, insulin, and leptin) promise to lead to new innovations in manipulation of growth and stature.

In 1985, the first case of Creutzfeldt-Jakob disease (CJD), a rare and fatal spongiform encephalopathy, was recognized in patients who had received GH derived from cadaveric pituitary glands; investigation disclosed that pituitary glands from which the GH was derived were contaminated with subviral particles (prions). Fortunately, the U.S. Food and Drug Administration (FDA) approved biosynthetic hGH the same year, with production of hGH by biological systems (Escherichia coli and, later, mammalian cells) inserted into the GH gene now yielding a virtually unlimited supply of hGH and eliminating risk of infection transmission. All currently available hGH is the 191-amino-acid recombinant human GH (rhGH) form, which is molecularly identical to endogenous GH.

Children with milder forms of inadequate GH secretion, previously excluded from GH therapy, can now be considered for treatment. Increased availability of hGH has also allowed investigation of its growth-promoting effects in poorly growing children who do not fit traditional definitions of GHD, many of whom were previously believed to be unresponsive to GH treatment. In addition, metabolic effects of GH apart from linear growth promotion are now being studied extensively, leading to new indications for and pharmacologic dosing of hGH therapy. The spectrum of disorders for which hGH has been prescribed and the number of children receiving treatment continue to increase (Table 2-5).

Concern about social and psychological harm of short stature, and hope for effective therapy, has resulted in increased referrals for growth-promoting therapy. However, data confirming that stature per se is a primary determinant of psychological health is limited, with only a minority of studies reporting detectable underachievement, behavior problems, and reduced social competency in short-statured children. Organic neuroendocrine dysfunction (eg, classic severe GHD), rather than stature itself, may correlate most closely with psychological and scholastic impairment. Though the physiologic benefits of hGH supplementation to children with severe GHD appear obvious, data confirming the efficacy of hGH therapy in improving the quality of life of non–GH-deficient recipients are scarce.

For many children, hGH treatment will be appropriate therapy after the cause of the growth problem, the concerns of patients and parents, and likelihood of success have been assessed. For most short children, however, efforts to build self-esteem through parental support, judicious selection of activities, and counseling may be more effective than hGH injections. The decision to institute long-term hGH therapy should include both careful physical and psychological evaluation to determine whether the degree of disability and likelihood of therapeutic benefit justify investment of the required emotional and monetary resources and potential adverse effects.

Intrinsic Short Stature

Many children are born with intrinsic traits that predispose them to short stature despite normal endocrine systems. These factors affect either the tempo at which growth is accomplished (constitutional growth delay) or the absolute growth potential of stature-increasing growth centers (eg, children with short familial stature, born SGA, or with defined genetic syndromes or skeletal dysplasias). The most severely affected children frequently demonstrate a combination of both traits. Standard laboratory evaluation usually reveals laboratory evidence for normal GH secretion and IGF-1 levels. However, recent discoveries (eg, mutation in the STAT5b component of GH signaling pathway) are beginning to provide physiologic explanations in a small number of cases of “normal” short stature. In addition, apparent normality of the GH/IGF-1 pathway does not preclude responsiveness to hGH therapy, and several causes of intrinsic short stature are now included in approved indications for hGH therapy.

Constitutional delay of growth and puberty

Constitutional delay of growth and puberty (CDGP) refers to children with short stature who have delay in the tempo of growth and pubertal maturation. Children who are short due to CDGP usually show deviation from growth along normal percentiles between 12 and 36 months of age and, by 3 years of age, begin to display normal growth velocity for age with height below their MPH percentile and parallel to, but usually below, the 5th percentile (although at a higher percentile when parents are very tall). The slower tempo of growth and maturation leads to late onset of puberty, exaggerated slowing of the prepubertal growth rate, and normal growth acceleration, but slight attenuation of overall growth accomplished during puberty. Adult height is usually within the normal range because of the longer period of growth prior to epiphyseal fusion, but usually in the lower part of the MPH range. By definition, children with pure CDGP should have BAs sufficiently delayed to result in normal predicted adult heights. Marked delay in pubertal growth can adversely affect spine mineralization, and osteopenia in men with a history of CDGP suggests that bone mineral accretion may be an intrinsic defect in those with CDGP.

The growth pattern typical for CDGP likely reflects secondary growth-slowing effects of exaggerated early-life inhibition of hypothalamic-pituitary-gonadal activity and delayed reawakening of this same system to trigger pubertal onset rather than primary abnormalities of the GH/IGF-1 axis. During early-to-mid-childhood, children with CDGP demonstrate normal GH secretion in response to provocative tests and have normal to slightly low serum IGF-1 and normal IGFBP-3 concentrations corrected for BA. However, in late childhood, children with severe CDGP often demonstrate low GH secretion in response to stimuli, suggesting transient GHD corresponding to the observed pubertal delay and growth deceleration resulting from low gonadal steroid production. Thus, if this physiologic phenomenon is not appreciated (ie, gonadal steroids are not administered before GH provocative testing), the diagnosis of true (rather than physiologically transient) GHD is suggested and treatment often initiated. The fact that approximately 70% of postpubertal adolescents diagnosed with idiopathic IGHD during childhood show restoration of normal GH secretion suggests that hGH treatment is frequently being used to treat patients with CDGP rather those who are truly GH-deficient.

Idiopathic short stature

Definition.

Idiopathic short stature (ISS) is a diagnosis of exclusion applied to otherwise healthy short children with no identified etiology for poor growth; that is, no systemic illness, endocrine disorders, genetic syndromes, bone dysplasias, or low birth weight and other factors compromising growth, such as depression or psychosocial deprivation, are identified. By definition, these children have heights less than 2 SD below the mean for a given age, sex, and, if available for comparison, specific population group.¹⁸

Diagnosis, etiologies, and natural history.

ISS can be defined by a height less than 5th percentile, growth velocity that is parallel to, but below, the normal growth curve, and BA congruent with chronological age. When there is no concurrent constitutional delay, time of pubertal onset is normal with a projected final height near the current growth percentile. The height of both parents is often less than 10th percentile with the calculation of the MPH and suggesting a final height that is reflective of that of the parents. As mentioned previously, a marked difference between parental height percentiles can compromise the utility of the MPH calculation. Though there is a tendency for a child’s height to trend toward the MPH, some will clearly follow one parent’s pattern closer than that of the other. Comparison with height percentiles of siblings may be helpful in this situation.

Some cases of familial or genetic short stature are likely to represent heritable, subtle disruptions in the GH/IGF-1/growth plate axis. An example of such a defect may involve receptor and postreceptor signaling abnormalities that create reduced linear growth accomplished in response to normal GH secretion. For instance, heterozygous GHR mutations are found in a small minority of children with ISS and poor response to GH, implicating receptor resistance to GH action. In addition, mutations in the SHOX (short stature homeobox-containing) gene region¹⁹ and heterozygous mutations in the IGF-1–associated ALS account for poor growth in another small subset of children with ISS. Given the complexity of possible genetic influences on the growth axis, the simple diagnosis of familial or genetic short stature will become increasingly inadequate to determine which children may warrant further evaluation and/or intervention.

Treatment.

Treatment of ISS with hGH is vigorously debated for economic and ethical reasons, as it is still unresolved as to what degree of shortness should be considered a normal variant and in what circumstances treatment is justified for a characteristic that is not a definable disease or disorder. Thus, the decision to medically intervene for those who meet the criteria for ISS is individually based on the degree of current childhood shortness, likelihood for an acceptable adult height without intervention, psychological stress associated with short stature, and evidence that treatment will have a positive effect on growth and quality of life. The FDA has approved hGH for treatment of ISS if the height is less than or equal to −2.25 SD below the mean (∼1st percentile), the predicted adult height falls below the normal range (160 cm [5 ft 3 in] for males, and 150 [4 ft 11 in] for females), epiphyses are open, and other causes of short stature are excluded. The average height velocity in the first year of treatment typically increases to 8 to 9 cm per year compared to 4.5 cm per year before treatment. However, the effect on ultimate height gained is relatively modest, estimated to be approximately 1 cm per year of hGH treatment.²⁰

Long-term hGH treatment of non–GH-deficient short children with ISS at recommended doses (0.375 mg/kg/week) can lead to statistically significant increases in final height in some children. As adult height is predominantly determined by the 85% of growth occurring before puberty, effective hGH therapy requires substantial growth acceleration prior to puberty and earlier treatment enhances outcome. Overall, girls appear to derive less height-increasing benefit, most likely related to later institution of treatment and earlier onset of puberty. Greater BA delay and taller MPH predict greater response, whereas low MPH predicts a poorer response; otherwise, there are presently no clinical (eg, pretreatment growth rate) or biochemical determinants that reliably predict long-term response to hGH therapy.¹⁸ Future studies may discover in the ISS population distinct genetic disturbances or subtle forms of GHD that respond differently to treatment. In the meantime, whether meaningful improvements in final height are sufficient to justify cost and commitment to several years of hGH therapy is still debatable.

Therapies that delay puberty (GnRH agonists = GnRHa) and/or prevent bone maturation (aromatase inhibitors) have also been used to increase height of children with ISS either on their own or in conjunction with hGH therapy. Pubertal suppression with GnRHa therapy is theoretically attractive, and it does not appear to adversely affect BMI, bone mineral density (BMD), body composition, and ovarian function, at least in late adolescence and early adulthood²¹; however, such therapy tends to decrease growth velocity to values below the normal prepubertal pace. Combined hGH/GnRHa treatment for 3 years may modestly improve adult height (mean ∼5 cm judged by the difference between realized adult height and initial predicted and target height).²² These data, coupled with the expense of GnRHa therapy, and the fact that pubertal delay is frequently a psychological concern for the short young adolescent, have caused enthusiasm for this approach to wane. Aromatase inhibitors (eg, letrozole and anastrazole), which markedly reduce conversion of androgen to estrogen and, therefore, limit estrogen-induced growth plate closure in both sexes, have also been used. In adolescent males, reports of 3 to 5 cm gain in near-adult height have been reported when these drugs are used in conjunction with testosterone therapy in males with CDGP or with GH therapy in males with GHD. Although aromatase inhibitors appear effective in increasing adult height of boys with short stature and/or pubertal delay, safety concerns, including vertebral deformities, a decrease in serum HDL-cholesterol levels, and an increase of erythrocytosis, are reasons for caution and preclude the use of such therapy outside a research setting. Future studies should clarify the impact of aromatase inhibition on BMD and bone quality in different compartments of bone and on vertebral body growth and strength. Other safety issues to be addressed include insulin sensitivity, maturing spermatogenesis, cognitive functions, vascular wall function, and prostate growth.²³

Turner syndrome

Definition and etiology.

Turner syndrome (TS) is the result of a haplotype insufficiency of the X chromosome with approximately 50% of the diagnosis being the result of a 45,X karyotype and the remainder resulting from mosaicism and X chromosome structural abnormalities. While relatively common in live-born females (∼1/2500), most affected fetuses are spontaneously aborted with those having a more mosaic pattern also having the greatest chances of survival.

Diagnosis and natural history.

Clinical appearance of TS is highly variable (Figure 2-17). At birth, congenital lymphedema, low hairline, nuchal skin, and pterygium coli (webbed neck) may be present. Other features include small jaw, high-arched palate, ptosis, downslanting palpebral fissures, epicanthal folds, midface hypoplasia, prominent ears, dysplastic nails, broad chest with hypoplastic nipples, cubitus valgus, multiple pigmented nevi, and short fourth metacarpals. Difficulties with recurrent otitis media may lead to conductive hearing loss, while sensorineural hearing loss of unknown etiology is common in adults. Intelligence is within the normal range, although difficulties with visual-spatial organization and attention deficit are common. Autoimmune disorders, especially thyroiditis and (to a lesser extent) celiac disease, occur with sufficient frequency to warrant screening.

A high percentage of TS patients will have cardiac, renal, and urologic abnormalities. Associated structural cardiac anomalies include elongated transverse aortic arch, nonstenotic bicuspid aortic valve, coarctation of the aorta, aortic dilation, and hypoplastic left heart syndrome. The urologic and renal defects include double collecting systems, horseshoe kidney, kidney rotational abnormalities, ureteropelvic and ureterovesical junction obstruction, and absence of one kidney. All children with the diagnosis of TS should have both cardiac and renal imaging.

Approximately 95% to 100% of girls with TS experience short stature with an average growth velocity in childhood of 4.44 cm for each year of BA advancement. Specialized growth curves for girls with TS are available for download (www.magicfoundation.org). The cause of growth failure and short stature is likely related to the underlying mild bone dysplasia affecting limbs disproportionately, abnormal body segment ratios (U/L ratio significantly above normal for age), and estrogen deficiency with (if untreated) lack of increased GH secretion during adolescence. Haploinsufficiency for the SHOX gene located on the long arm of the X chromosome is a key contributor to growth restriction. Evidence of growth failure may be present at birth, but typically birth length is in the lower range of normal. Untreated mean final height, however, is approximately 20 cm below the female average (Figure 2-18).

Treatment.

Although girls with TS have normal GH responses to provocative testing, abnormalities in growth plate cartilage rendering resistance to GH, and relatively low GH secretion after 8 years of age are reported as potential GH/IGF-1 axis contributors to poor growth. hGH therapy can increase growth rate and height achieved in TS and is an FDA-approved indication. Standard treatment uses a dose of approximately 0.35 mg/kg/week administered by daily subcutaneous injections as soon as a girl with TS drops below the 5th percentile of the normal growth curve (as early as 2 years of age). In a randomized, controlled clinical trial, the growth failure, typical of early childhood in girls with TS, was corrected when GH treatment was initiated by four years of age,²⁴ suggesting that, the younger the patient is at hGH initiation, the smaller the height deficit to be bridged and the faster that height is normalized. Long-term follow-up of a Canadian controlled study showed that the treatment group achieved a mean increase of 7.1 cm in height and height, on average, 5.1 cm closer to MPH than did the control group.²⁵ Insulin sensitivity decreases during hGH therapy in girls with TS and remains altered 5 years after treatment is stopped; it remains uncertain whether this insulin resistance is a posttreatment adverse effect of hGH or part of the natural history of TS. In girls older than 8 years with marked short stature treated with hGH, concomitant administration, until growth completion, of oxandrolone (∼1.25-2.5 mg/day), an anabolic steroid not aromatized to estrogen, appears to add 2 to 4 cm to adult height (Figure 2-19).^26,27 Possible side effects of oxandrolone—clitoral enlargement, glucose intolerance, and elevation in liver enzymes—are dose-dependent and extremely rare with lower dosages (eg, 1.25 mg/day). Thus, an adult height above the lower limit of normal for American women (150 cm) is now an attainable goal for many girls with TS.

Ovarian failure is another major issue for X haploinsufficiency, with either complete (∼75%) or partial (∼25%) absence of estrogen-induced sex characteristics (ie, breast development and menses). Streak gonads can be seen in two-thirds of patients and one-third have below normal ovarian volumes. Ovarian function is related to underlying karyotype; those who achieve menarche or, very rarely, pregnancy are predominately those with mosaic karyotypes. Evidence of gonadal failure can be seen in the newborn period when gonadotropins are elevated until approximately 4 years of age when they decline to normal childhood level and then rise again to high levels at the usual age of pubertal onset. Estrogen replacement is necessary for those girls with complete ovarian failure or for those who demonstrate mid-pubertal arrest. Timing of estrogen treatment is individualized to maximize height attainment, psychological well-being, and bone mineralization. With oral estrogen therapy, there is a clear relationship between earlier treatment and reduced height achieved with GH therapy. However, awareness of likely hGH response-inhibiting effects of orally administered estrogens (due to concentration-dependent inhibition of hepatic IGF-1 production) and the advent of effective low-dose transdermal estrogen preparations may allow more age-appropriate sex hormone replacement without compromise of growth potential. In one double-blind, placebo-controlled trial, it was shown that combining childhood ultra–low-dose estrogen with hGH improved growth rate, adult height attainment, and provided other potential benefits associated with early initiation of estrogen replacement, suggesting synergism between hGH and low-dose estrogen in promoting growth.²⁸ Thus, initiation of treatment of low-dose estradiol at an average age of 9.3 years to girls with TS treated with hGH may be a better option than the common practice of delaying estrogen-replacement therapy until the mid-teens because of the perceived possibility of interfering with growth.

Short stature homeobox containing gene haploinsufficiency

An important cause of short stature is short stature homeobox (SHOX) gene haploinsufficiency, affecting the development of the extremities. SHOX mutations occur with an estimated incidence of roughly 1 in 1000 newborns and, therefore, is one of the most common genetic defects associated with growth failure and skeletal deformities.²⁹ Heterozygous mutations of SHOX are detected in 50% to 90% of patients with Leri-Weill dyschondrosteosis, in the vast majority of girls with TS, in 2% to 15% of patients with ISS, and in patients with Madelung deformity (bayonet deformity of the wrist). Homozygous mutations of the SHOX gene have been identified in patients with Langer mesomelic dysplasia.

The SHOX gene is located on the pseudoautosomal region 1 (PAR 1) at the distal end of the short arms of the sex chromosomes at Xp22 and Yp11.3. This region contains genes that escape X inactivation and, therefore, SHOX is expressed on both sex chromosomes. SHOX plays a major role in limb development by acting as a promoter for linear growth and as a repressor for growth plate fusion and skeletal maturation in the distal limbs. SHOX haploinsufficiency may account for the premature terminal differentiation of the proliferative chondrocytes with progression to the hypertrophic phenotype and accelerated early growth plate fusion of the distal limb bones, resulting in compromised linear growth. These skeletal lesions become evident with puberty and females are more severely affected than males.

Mutations or deletions of SHOX are associated with a wide range of phenotypic expression, including short arms in 92%, bilateral Madelung deformity in 73%, short stature in 54%, cubitus valgus, and high-arched palate. In addition, muscular hypertrophy of the calves is found in one-third of affected individuals. Radiographic features of SHOX deficiency include short metacarpals/metatarsals with metaphyseal flaring, triangularization of the radial head, and radial/tibial bowing. The absence of any or all of these signs, including Madelung deformity, does not exclude SHOX haploinsufficiency.

Treatment.

GH treatment has been shown to be effective in increasing linear growth in SHOX haploinsufficiency. Patients show an excellent growth spurt; however, growth of the lower extremities is less robust than that of the trunk and arms, indicating that rhGH was effective for height gain, but with a tendency to induce disproportionate growth as well. Concomitant use of a GnRH analog may improve adult height gain in patients with isolated SHOX defects.³⁰

Noonan syndrome

Definition, diagnosis, and natural history.

Noonan syndrome (NS) is an autosomal dominant disorder occurring in both sexes. It is characterized by short stature (mean height 2 SD below population mean, usually in the absence of GHD), webbing of the neck, facial dysmorphology (low posterior hairline, ptosis, and ear malformations) (Figure 2-20), and cardiovascular anomalies. In contrast to left-sided heart defects typical for TS, heart defects in NS are primarily right-sided (pulmonic and peripheral pulmonic stenosis). Gonadal function is usually preserved, although small genitalia and cryptorchidism occur with increased frequency in males, and delays and incomplete progression in puberty can occur in both sexes. Mental retardation of variable degrees is observed in 25% to 50% of children with NS. Mutations that cause NS alter genes encoding proteins with roles in the RAS-mitogen–activated protein kinases (MAPK) signal transduction pathway, which is implicated in several developmental processes controlling morphologic determination, organogenesis, synaptic plasticity, and growth.³¹ So far, heterozygous mutations in nine genes (PTPN11, SOS1, KRAS, NRAS, RAF1, BRAF, SHOC2, MEK1, and CBL) have been documented to underlie this disorder or clinically related phenotypes. Based on these recent discoveries, the diagnosis can now be confirmed molecularly in approximately 75% of affected individuals. The PTPN11 (protein tyrosine phosphatase, nonreceptor type 11) mutation is the most common mutation found in this clinical syndrome (in ∼50% of cases).

Birth weights are typically normal, but infancy and childhood growth diverges from normal percentiles. Abnormal body proportions may be present, and pubertal growth rates are attenuated so that patients reach median adult heights of 162.5 cm and 152.7 for men and women, respectively. The precise etiology of poor growth remains unknown. GH secretion abnormalities do not explain the poor growth, although occasional reductions in GH production are reported.

Treatment.

Treatment of children with NS and short stature is the most recent FDA-approved indication for hGH therapy. Results from a 4-year analysis of GH therapy in NS subjects demonstrated an improvement in height SDS from mean –2.65 ± 0.73 at baseline to –1.32 ± 1.11, with sustained growth over multiple years of treatment. The mean increase (between 9-10 cm) was similar for boys and girls.³² Some studies report different response to hGH treatment based on the genotype of NS, but others have found no correlation.

Skeletal dysplasias (achondroplasia and hypochondroplasia)

Definition and diagnosis.

Skeletal dysplasias (osteochondrodysplasias) are a heterogeneous group of genetically transmitted bone and cartilage abnormalities that result in orthopedic complications and varying degrees of short stature with disproportionate size and/or shape of limbs, skull, and spine. There are over 400 osteochondrodysplasia conditions identified on the basis of the physical examination, radiologic findings, and molecular genetics. Abnormalities can predominantly involve epiphyseal, metaphyseal, or diaphyseal bone growth. Further characterizations are made based on biochemical and genetic studies, but exact diagnosis is still often difficult. Body proportion measurement is critical, including arm span, sitting height, U/L body segment, and head circumference. Radiologic studies clarify involvement of the long bones, skull, and vertebrae. Although each skeletal dysplasia is relatively rare, collectively their incidence is 1/5000.

Etiologies, diagnosis, and treatment.

Two of the more common osteochondrodysplasias are achondroplasia and hypochondroplasia. Both result from activating mutations in the fibroblast growth factor receptor 3 (FGFR3) gene, but in different domains. The FGFR3 gene is located on the short arm of chromosome 4 and normally functions to negatively regulate bone growth. The spectrum of these disorders is caused by different gain-of-function mutations in the FGFR3 gene, which enhance growth-inhibiting effects through various downstream signaling pathways.

Achondroplasia occurs in approximately 1:20,000 live births. While inherited in an autosomal dominant pattern, most new cases are new, spontaneous mutations. Homozygous children usually die in infancy owing to restrictive lung disease. Essentially all patients with the classic features of achondroplasia have the same glycine to arginine substitution at position 380 (G380R), which maps to the transmembrane domain of the receptor. The negative regulatory influence of FGFR3 on the growing skeleton is exerted mainly in the growth phase, in which it reduces the rate of cartilage template formation and turnover necessary for bone elongation. FGFR3 inhibits both the proliferation and terminal differentiation of growth plate chondrocytes and synthesis of extracellular matrix by these cells.

Skeletal features include megalocephaly, low nasal bridge, lumbar lordosis, short trident hands, rhizomelia (shortness of the proximal legs and arms) with skin redundancy, and small iliac wings. The spinal abnormalities are small cuboid-shaped vertebral bodies with short pedicles and narrowing of the lumbar interpedicular distance. A small foramen magnum creates risk for hydrocephalus, whereas kyphosis and spinal stenosis may lead to spinal cord compression. Skeletal dysmorphology and abnormal growth velocity are present from infancy, although short stature may not be evident until after 2 years of age. Mean adult heights in males and females are 130 and 125 cm, respectively. Growth curves specifically for achondroplasia have been developed and are of value in following patients.

Patients with achondroplasia demonstrate normal GH secretion, IGF-1 levels, and IGF1R activity. Reports of response to hGH are variable both in terms of the improvement in linear growth and velocity, but also in the degree of improved ratio of lower limb length. Though there is some evidence that growth can be accelerated in the short term, there is no evidence of significant increases in final adult stature following the administrationof hGH.³³ Lower limb-lengthening is an operative treatment aimed to achieve physiologic proportions in the body and improvement of height. Lengthening procedures include either tibial only, femoral only, or both tibial and femoral lengthening. The average lengthening gains reported range between 5.7 and 20.5 cm. A number of international research projects are currently underway to investigate potential therapeutic strategies, aimed at reducing excessive FGFR3 output. They include strategies to interfere with FGFR3 synthesis, block its activation, inhibit its tyrosine kinase activity, promote its degradation, and antagonize its downstream signals.

Hypochondroplasia is an autosomal dominant disorder, previously described as a “mild form” of achondroplasia, but the two disorders do not occur in the same family. Approximately 70% of affected individuals are heterozygous for a mutation in FGFR3 gene, but locus heterogeneity exists as other unidentified mutations cause a very similar phenotype.

In patients with hypochondroplasia, the facial features of achondroplasia are absent, and both short stature and rhizomelia are less pronounced. Poor growth may not be evident until after 2 years of age, but stature then deviates progressively from normal. Because of the more subtle changes compared to those seen in achondroplasia, the diagnosis of hypochondroplasia can be exceedingly difficult to make in young children. Mild variants of the syndrome may be difficult todistinguish from normal without careful measurement of U/L ratios and consultative radiologic opinions. Occasionally, severe short stature and body disproportion are not appreciated until near-adulthood when, following reduced growth achieved during the pubertal growth spurt, these features become more apparent. Adult heights typically are in the 120- to 150-cm range. Outward bowing of the legs may be accompanied by genu varum, diminished lumbar interpedicular distances between L1 and L5, and there may be flaring of the pelvis and narrow sciatic notches.

Modest improvement in PAH for those receiving hGH therapy is reported, but these estimates may be exaggerated by rapid growth during the first year of treatment. Proportionality of the growth response (ie, spine vs extremity) to hGH may be dependent on the specific genotype of the IGF-1 gene locus. As in other conditions of intrinsic shortness, it is possible that higher hGH dosage, concomitant GnRH analog treatment, and/or use of aromatase inhibitors could alter final height in these patients. It is likely that some mildly affected (and undiagnosed) children with hypochondroplasia are receiving hGH treatment within the approved ISS indication. More severely affected patients would require leg-lengthening procedures to achieve a height in the normal range.

Prader-Willi syndrome

Definition, etiology, and natural history.

Prader-Willi syndrome (PWS) results from deletion of the paternal allele, maternal disomy, or abnormal methylation of a critical region of chromosome 15 (15q11-13). Affected children are characterized by obesity, hypotonia, short stature, small hands and feet, hypogonadism (Figure 2-21), and behavioral abnormalities. Many features of PWS reflect hypothalamic dysfunction, including hyperphagia, sleep disorders, deficient GH secretion, and hypogonadism. Interestingly, ghrelin levels in children with PWS are elevated compared with BMI-matched obese controls and, thus, may play a role as an orexigenic factor driving the insatiable appetite and obesity found in PWS. With an incidence of 1 in every 10,000 to 12,000 births, PWS is the most common genetic syndrome associated with marked obesity.

Children with PWS display moderate intrauterine (average −1 SDS) and postnatal growth delay. Usually after age 2 or 3 years, when caloric intake increases and obesity begins to develop, growth rates improve. In contrast to the normal or accelerated growth typically seen in healthy non-PWS obese children, the growth rates are borderline normal or diminished in the midst of excessive weight gain. Standardized growth curves for weight, length, head circumference, weight/length, and BMI for non–GH-treated infants with PWS between 0 and 36 months of age are available.³⁴ In addition, the lack of normal pubertal growth usually results in reduced adult stature (mean 152 cm for adult PWS males and 146 cm for adult PWS females). GH responses to provocative testing are low-normal or blunted in PWS. Levels of IGF-1 are relatively low (mean ∼ −1.5 SDS) compared to normal-weight age-matched children, but not as low as in those with severe GHD. This likely reflects a stimulating influence of nutritional energy excess. Insulin levels are lower in children with PWS than in “healthy” obese children and rise to the levels of obese children only with hGH-replacement therapy, suggesting relatively pretreatment-heightened insulin sensitivity compatible with reduced GH secretion and accentuated deposition of subcutaneous (rather than visceral) fat.

Treatment.

Growth failure related to PWS is an FDA-approved indication for hGH treatment in the United States, even in the absence of demonstrable GHD. The approved indication in Europe includes PWS-associated abnormalities of body composition. In a recent expert consensus meeting, it was recommended that GH stimulation testing should not be required as part of the therapeutic decision-making process in infants and children with PWS.³⁵ Because serum IGF-I is a useful biomarker for monitoring compliance with treatment as well as sensitivity to hGH, baseline IGF-I levels should be determined. hGH treatment (at the recommended dosage of ∼1 mg/m²/day or 0.24 mg/kg/week) leads to a first-year growth rate of 10.1 ± 2.5 cm/year, a rate equal to that of other children with severe GHD, with a second-year decrease in growth rate to 6.8 ± 2.3 cm, corresponding to a mean growth velocity SDS of 2.2 ± 2.2. Maintenance of catch-up growth rates in subsequent years often requires higher dosage of hGH (eg, 1.5 mg/m²/day) leading to a cumulative change in height SDS during treatment of 1.8 ± 0.6.³⁶ Mean increases in adult height of 10 cm and 6.5 cm in hGH-treated males and female PWS patients, respectively, have also been reported.

Patients with PWS also appear to benefit from the metabolic effects of hGH therapy. Pretreatment body composition studies of children with PWS (using the DXA method) have revealed a markedly increased percent of body fat (mean body fat 45%-49%) and low lean body mass; hGH therapy leads to significant reductions in body fat (8%-10%) and increased lean body mass (see Figure 2-21). With more prolonged treatment given at doses up to 1.5 mg/m²/day, body composition, physical strength, activity, stamina, coordination ability, pulmonary function, and sleep quality continue to improve, but function in all parameters remains below normal. This likely reflects the influence of non-GH factors regulating body composition affected by the genetic mutation causing PWS and/or the relatively late institution of hGH therapy. Recent studies indicate that earlier institution of hGH therapy during infancy favorably alters the natural history of PWS regarding height, body composition, lipid levels, and perhaps cognitive function.³⁷

Exclusion criteria for starting hGH in patients with PWS include severe obesity, uncontrolled diabetes, untreated severe obstructive sleep apnea, active cancer, and active psychosis. Scoliosis should not be considered a contraindication to hGH treatment in patients with PWS. Children with PWS have a high incidence of both central apnea and obstructive apnea. Families should be informed about the potential association between rhGH therapy and sleep apnea and unexpected death. In fact, polysomnography performed before starting hGH therapy is recommended. Because hGH therapy can theoretically lead to lymphoid tissue growth in children, hGH therapy is contraindicated in children with breathing difficulties until ear-nose-throat (ENT) evaluation has been done and corrective therapy instituted when indicated. Therapy should not be initiated during an acute respiratory infection, but it need not be interrupted during subsequent episodes of respiratory infection unless indicated because of the onset of breathing difficulties.

Another potential adverse effect of hGH therapy of particular concern in PWS patients includes changes in glucose tolerance. However, fasting glucose, insulin, and HbA1c levels do not generally change during prolonged hGH therapy in PWS patients. Rarely, symptoms consistent with pseudotumor cerebri occur and resolve with temporary withdrawal followed by dose-graduated reintroduction of hGH therapy. Baseline scoliosis screening and annual assessments during treatment have shown no differences between treated or observed PWS patients.

Intrauterine growth restriction or small for gestational age

Definition and etiology.

Children born small for gestational age (SGA) are defined as having a birth weight and/or length less than 2 SDS below the mean for their gestational age and sex (below the 3rd percentile). Accurate classification of infants who are SGA depends on access to the appropriate reference data for birth weight and birth length available for specific populations for identifying those at risk. In the United States, the most commonly used data on intrauterine growth come from charts developed by Usher and McLean.³⁸ In more than 80% of infants who are born SGA, catch-up growth occurs during the first 6 months of life.

Although many of these children achieve a normal growth pattern, approximately 10% of children born SGA will remain −2 SD or less for height throughout childhood, adolescence, and into adulthood. Those who fail to show linear catch-up growth by 2 to 3 years of age comprise a substantial population of children with small stature. The mean adult height in SGA babies is reduced by a mean 3 to 4 cm compared to their MPH and the severity of this effect is proportional to the extent of fetal impairment.

Diagnosis.

Tests of the GH-IGF-1 axis are usually normal. Low levels of IGF-1 and IGFBP-3, suspected in utero secondary to fetal undernutrition, may persist and contribute to postnatal growth retardation, although levels of these proteins do not correlate with postnatal growth. Relative insensitivity to GH and IGF-1 is also possible, as elevations in GH levels before and IGF-1 levels after hGH therapy are noted in some SGA infants. Subsequent endocrine disorders associated with being born SGA include premature adrenarche, ovarian hyperandrogenism, and reduced pubertal growth. A propensity for faster tempo in puberty and, thus, a quicker rate of growth plate maturation and fusion can further jeopardize adult height potential.³⁹

Natural history of disordered fetal growth.

In the last two decades, there has been increasing evidence of important consequences of disordered fetal growth not only on childhood and adult stature but also on later health status, a phenomenon referred to as the “fetal origins of adult disease.” In particular, interest has focused on the increased risk that individuals born with impaired fetal growth have for development of obesity, insulin resistance, polycystic ovarian syndrome, type 2 DM, hypertension, and stroke.⁴⁰ Intrauterine growth restriction (IUGR) during a critical developmental period of heightened sensitivity to the nutrient environment appears to lead to permanent alterations in the function of hormonal and regulatory pathways, as well as structural changes in organs. These regulatory and structural changes, in the setting of later nutrient excess, lead to increased risk of diseases related to reduced insulin action.

The association between birth weight and later cardiovascular disease was initially thought to reflect alterations in structure due to nutrient deficiency. However, subsequent cross-sectional and longitudinal studies demonstrated a critical interaction between birth weight and postnatal growth, with excessive weight gain after birth being a primary determinant of risk for later cardiovascular disease. In overweight adolescents, for example, the prevalence of the cluster of insulin resistance–related disorders, often called the metabolic syndrome, is substantially higher in those with a low birth weight. Though the prevalence of metabolic syndrome is also higher in normal-weight children born with low weight than their normal birth weight peers, the difference is substantially smaller, suggesting an important contribution of later weight gain. Furthermore, weight gain in early childhood appears to be of particular importance in this phenomenon; young adults with a history of low birth weight who gained weight rapidly during the first year of life are more likely to have reduced insulin sensitivity and secretion, higher centrally distributed body fat, hypertension, and dyslipidemia (low HDL and elevated triglycerides). Weight gain during the first 3 to 6 months of life appears to be of particular importance.

The mechanisms of these phenomena are only beginning to be clarified and likely involve a combination of alterations in structure (eg, changes in pancreatic β-cell mass), function (eg, insulin signaling in liver, adipose tissue, and muscle), traditional genetics, and epigenetics. However, the literature to date is clear that both intrauterine nutrient restriction and postnatal weight gain interact to promote increased risk of insulin resistance, deleterious body composition, diabetes, and cardiovascular disease. Thus, those children who are born with the lowest weight and have the greatest postnatal weight gain are at the highest risk of later health consequences, whereas normal-weight infants with the lowest postnatal weight gain have the lowest risk. Normal-weight infants with rapid postnatal weight gain and low-weight infants with slow postnatal weight gain have intermediate risk. Importantly, long-term hGH treatment does not appear to increase risk for type 2 DM and metabolic syndrome.⁴¹ Metformin treatment of some female children born SGA may mitigate severity of metabolic syndrome, delay menarche, and improve height prognosis.

On the other hand, the relationship between birth weight and later risk of obesity, diabetes, and cardiovascular disease is U-shaped, suggesting that fetal overnutrition is also a risk for adult disease,⁴² though this phenomenon is less well understood than is intrauterine restriction. The infant born to a pregnancy complicated by diabetes has substantially increased risk of early-onset obesity, metabolic syndrome abnormalities, and both childhood- and adult-onset type 2 DM. Indeed, being born to a diabetic pregnancy is the single strongest risk factor for development of type 2 DM prior to age 18, accounting for over 40% of the risk in most studies. Such infants are heavier at birth, and have higher BMI, waist circumference, visceral and subcutaneous adipose tissue, and a more centralized fat distribution in later childhood and adolescence than do the offspring of prediabetic or nondiabetic women. Furthermore, children exposed to maternal diabetes have a faster BMI growth trajectory beginning at approximately 2 years of age. However, the increased risk for diabetes is only partially accounted for by the development of obesity in these offspring. Studies have suggested alterations in both β-cell mass and function, as well as insulin sensitivity in the offspring of diabetic mothers, but evidence to date is limited and no comprehensive model has been developed. Finally, children born to pregnancies complicated by maternal obesity without DM also have increased risk for early-onset obesity and cardiovascular disease, suggesting that fetal overnutrition short of overt maternal DM is deleterious to offspring. Direct associations between maternal fasting glucose and risk of childhood obesity and diabetes have been reported, as have associations with maternal free fatty acids (FFAs) and triglycerides in the face of normal glucose metabolism. Further study of longitudinal cohorts of children born to obese mothers will be needed to better understand the role of fetal overnutrition on risk for adult disease.

Treatment.

Results of therapeutic trials of hGH therapy for children born SGA with persistent growth retardation at 2 years of age have been sufficiently encouraging to justify FDA approval of hGH therapy, with a dosing range up to (clearly supraphysiologic) 0.48 mg/kg/week for this indication. Novel use of intermittent high-dose hGH therapy has been successful for SGA patients and could be potentially applied to other hGH treatment indications; however, long-term studies suggest that additional height gained with either intermittent or high-dose hGH treatment is modest (∼0.5 SD or 1.25 in) compared to sustained lower-dose treatment. Whether children born SGA derive substantial psychological benefits from enhanced childhood stature (even without enhanced adult stature) remains uncertain. However, there is some evidence that intelligence and psychosocial functioning may be enhanced for thoseSGA children on hGH therapy.³⁹ Given the aforementioned propensity of insulin resistance in this population, exacerbating this metabolic dysfunction has been a concern. Although higher-dose hGH therapy has increased insulin levels in some of these children, clinically significant deterioration in glucose homeostasis has, thus far, not been identified. Auxiliary height-enhancing treatment for short children born SGA may include administration of GnRHa if accelerated pubertal tempo is observed. Similar to ISS, aromatase inhibition may prove useful in slowing growth plate maturation, but it is still considered experimental.

Russell-Silver syndrome

Definition, diagnosis, and natural history.

Russell-Silver syndrome (RSS) is characterized by IUGR, poor postnatal growth, body asymmetry, relative macrocephaly, triangular face, asymmetry, and feeding difficulties (Figure 2-22). Height outcome can also be compromised by early onset of puberty and/or accelerated progression through puberty. Two main molecular mechanisms are known to be involved: maternal uniparental disomy for chromosome 7 (mUPD7) (∼10%-15%) and methylation abnormalities of chromosome 11p15 (∼50%, likely due to suppressed IGF-2 production).

Treatment.

Children with RSS are often treated with hGH because of SGA, and particularly in children with hypoglycemia. There are, however, no randomized controlled trials and only a limited number of reports on the effectiveness of hGH treatment in RSS. Among 26 patients (16 males) with RSS who received treatment with hGH for 10 years and attained final height, a significant improvement of growth has been shown; the median height at the commencement of treatment was –2.7 SDS and increased to –1.3 SDS (p = .001). Predictors of adult height outcome were the height at the start of treatment and the height gain at onset of puberty.

Down syndrome

Definition, diagnosis, and natural history.

Short stature is a prominent feature of Down syndrome (DS; trisomy 21). For reasons that remain obscure, children with DS are born approximately 500 gram and 2 to 3 cm smaller in birth length than normal and continue postnatal growth at subnormal rates. Although subnormal GH and IGF-1 levels have been identified in some children with DS, most show normal stimulated GH levels. Skeletal maturation is typically delayed, and pubertal growth acceleration is blunted. Graphing a child with DS on normal growth chart alone may lead to erroneously attributing declining growth percentiles solely to the syndrome rather than pursuing other etiologies such as hypothyroidism or celiac disease which occur with increased frequency. However, since available specialized DS-specific growth charts likely no longer reflect the current population, it has been recently recommended that growth of these children be plotted on both the specialized charts and the CDC growth charts until new standards are developed.⁴ This will allow comparisons of growth to the general population of children and to the references for children identified with a given condition.

Treatment.

Though short stature is an expected feature of the natural history of DS, failure of a child with DS to demonstrate appropriate growth rate when plotted on the DS-specialized growth curve should prompt evaluation for attenuated growth as detailed previously. Given a strong association of DS with autoimmune thyroid disease, periodic evaluation of thyroid status as well as for celiac disease is indicated. Concern regarding increased risk in DS of leukemia and DM, coupled with uncertainty about benefits of treatment, has thus far limited enthusiasm for growth-promoting therapy, and hGH therapy for DS is not a FDA-approved indication.

Endocrine Causes of Disordered Growth

Endocrine systems that control GH secretion and its downstream messenger IGF-1, pubertal development, thyroid activity, the adrenal cortex, calcium metabolism, and glucose homeostasis all have effects on childhood growth. Consequently, a wide variety of endocrine disorders, either congenital or acquired and resulting either in deficiency or excessive production of hormone effect, are associated with disordered growth.

Deficiency in the growth-promoting action of the GH/IGF-1 axis can result from any defect in the pathway from synthesis, release, or receptor binding of GH through the generation of its downstream messenger IGF-1, its association with IGFBPs and receptor binding, and subsequent cell signaling pathways. Genes regulating and physiologic pathways involved in this axis are continually being discovered, improving understanding and creating new diagnoses and treatment possibilities. The multitude of factors involved in just this one pathway illuminates why current diagnostic tests of the GH/IGF-1 axis are frequently inadequate to fully characterize the cause of a child’s growth disorder.

IGF-1 deficiency

Definition.

As IGF-1 is a major mediator of skeletal growth, its deficiency can result in severe growth failure. Given the central role of IGF-1, growth failure can be classified based on IGF-1 levels as (1) primary IGF-1 deficiency: low IGF-1 levels in the presence of normal GH production, resulting from defects involving the GHR, its postreceptor signaling cascade, or the IGF-1 gene itself and (2) secondary IGF-1 deficiency: low IGF-1 levels due to a defect in GH production or secretion. GHD can either be isolated due to a GHRH or GHRH receptor gene deletion, a GH-1 mutation, or bioinactive GH, or can be part of a combined pituitary hormone deficiency complex resulting from mutations in one of several transcription factor genes that regulate the development of the anterior pituitary gland (see Table 2-4).

Thus, the IGF-1 deficiency syndrome describes a generic condition resulting from various etiologies that share similar, but not identical, clinical phenotypes. If GH or IGF-1 deficiency is acquired, clinical signs and symptoms will appear at a later age.

Natural history of congenital IGF-1 deficiency.

Severe IUGR is not part of secondary IGF-1 deficiency but is often seen in infants with primary IGF-1 deficiency of other causes, confirming (1) the critical role of IGF-1 in intrauterine growth and (2) the relative independence of IGF-1 production from GH regulation in utero. Most children born with secondary IGF-1 deficiency have a birth size that is normal or within 10% of normal, but it can be lower in severe congenital GHD or GHI. Breech deliveries and perinatal asphyxia are more common in infants with congenital GHD, and neonatal morbidity can include hypoglycemia and prolonged jaundice with direct hyperbilirubinemia caused by cholestasis (usually due to combined cortisol and GHD) and/or giant cell hepatitis. When GHD is combined with deficiency of adrenocorticotropic hormone (ACTH)/cortisol, hypoglycemia may be severe. The combination of GHD with gonadotropin deficiency leads to microphallus, cryptorchidism, and hypoplasia of the scrotum. GHD (or GHI) should, therefore, be considered in the differential diagnosis of neonatal hypoglycemia and of microphallus/cryptorchidism. Of note, microphallus can occur with IGHD.

Growth failure can occur immediately after birth, but not always, although by 6 to 12 months of age the growth rate invariably deviates downward from the lower part of the normal growth curve. Skeletal proportions tend to be relatively normal but correlate better with BA than with chronological age. BA is delayed and (in the absence of hypothyroidism) is similar to the height age. In acquired GHD, however, as from a CNS tumor, BA may approximate the chronological age. Weight/height ratios tend to be increased, and fat distribution is often “infantile” or “doll-like” (also referred to as “cherubic or angel-like”) in pattern. Musculature is poor, especially in infancy, and can cause delay in gross motor development. Fontanelle closure is often delayed, the voice infantile due to hypoplasia of the larynx, hair growth sparse and thin, and nail growth slow. Associated midline abnormalities may be present, including cleft palate and lip, single maxillary central incisor, and other mid-facial bony and soft-tissue abnormalities (Figure 2-23). Even with normal gonadotropin production, the penis is often small, and puberty is usually delayed. Adult height data in patients with untreated GHD are not plentiful, but a survey of reports indicates mean final height −3 to −4 SD. Anatomical abnormalities include dysgenesis of the pituitary stalk, ectopic placement of the posterior pituitary inferior to the median eminence, and diminished volume of the anterior pituitary.

Etiologies

1. Idiopathic and isolated GHD

   Most children receiving hGH are diagnosed as having either acquired, idiopathic, and/or isolated GHD (IGHD) (Figure 2-24). This diagnosis, however, should always be considered suspect, especially in the immediate prepubertal period owing to normal transient reduction in GH secretion at this time). Retesting of patients diagnosed with IGHD reveals normal GH secretion in more than 70% of those with “partial” GHD and normal magnetic resonance imaging (MRI) findings.⁴³ Additionally, IGHD patients with normal MRI findings have mean adult heights equal to those with ISS treated with GH, supporting the likely biological similarity of these two entities. In contrast, over 90% of patients with combined pituitary hormone deficiency, with or without structural abnormalities of the hypothalamic-pituitary area, demonstrate sustained GHD. Whether these results simply cast doubt upon the validity of the initial diagnosis or whether children with IGHD may truly normalize is not clear.

2. Acquired GHD

   The most concerning cause of acquired GHD is a tumor in the hypothalamic-pituitary region. The most common tumor that arises in this region in childhood is a craniopharyngioma. Craniopharyngioma develops from remnants of Rathke pouch and can originate anywhere along an axis extending from the sella turcica, through the pituitary stalk, to the hypothalamus. Tumors are invariably partly cystic and have a benign histologic appearance; however, as they enlarge, they invade nearby vital anatomical structures and, thus, can have an aggressive clinical course. The presenting symptoms are dependent both on the location of the tumor and on its size. The tumor may compress the pituitary gland and compromise vision by effects on the optic nerves. While classically associated with “bitemporal hemianopsia” (limited lateral vision), most typical are “quadrianopsias,” which are smaller, more irregular field cuts. The tumor may block the third ventricle and cause obstructive hydrocephalus, leading to headaches, vomiting, and blurry vision secondary to papilledema. Both the tumor and its treatment can lead to significant neurologic and endocrinologic complications. The optimal treatment approach for craniopharyngioma tumors has been subject to debate for many years. Though complete resection has been preferred, extensive surgeries often result in multiple long-term complications such as hypopituitarism; hypothalamic dysfunction; hypothalamic obesity; and neurocognitive impairments in areas of memory, learning, and school performance. During the past decade, less aggressive surgical approaches, such as limited resection or cyst decompression, sometimes accompanied by radiation therapy, have been adopted.⁴⁴ Other tumors that can cause GHD include germinoma, glioma/astrocytoma, and pituitary adenoma (rare prior to adulthood). Other causes of acquired hypopituitarism in childhood include previous brain infection (encephalitis and/or meningitis), hydrocephalus (even without an underlying tumor), infiltrative disorders (Langerhans cell histiocytosis, tuberculosis, sarcoidosis, and lymphocytic hypophysitis), vascular abnormalities, and traumatic brain injury.

3. GHD following cancer treatment

   Children who receive radiation to the hypothalamus/pituitary commonly develop IGF-1 deficiency secondary to GHD. This effect is dose-dependent: Above 18 Gy, the pubertal increase in spontaneous GH secretion is blunted; above 24 Gy, all spontaneous GH secretion is diminished; and above 27 Gy, GH secretion in response to provocative stimuli is reduced. The likelihood of GHD is increased by larger fraction size administered over shorter time intervals, younger age, and with increasing time following treatment. Growth failure is also observed commonly during the acute phase of chemotherapy, primarily because of poor nutrition, catabolic stress of illness, and growth-suppressing effects of adjunct medications (eg, glucocorticoids). The addition of chemotherapy to radiation exacerbates growth failure. Thus, brain tumor survivors should be monitored every 3 to 6 months for reduction in growth rate and assessment of upper and lower segment lengths.

   Other hormone deficiencies and direct effects on skeletal tissues can also adversely affect growth of children following radiation. Particularly at cumulative radiation doses exceeding 50 Gy, variable degrees of hypogonadism, along with thyroid-stimulating hormone (TSH) and ACTH deficiency, are likely to occur, with normal function in all pituitary hormone systems found in fewer than 10% of the subjects. Irradiation of skeletal tissues can arrest chondrogenesis at the epiphyseal growth plate, and spinal irradiation often results in loss of vertebral growth, scoliosis, development of abnormally low U/L body segment ratios, and, as a result, failure to show expected gains in height during puberty. These direct growth-inhibiting effects of radiation treatment are not corrected by hGH therapy.

4. Insensitivity to GH action

   There is increasing recognition of defects in the distal portion of the GH/IGF-1 axis that extend from the binding of GH to its receptor and to the postreceptor actions of IGF-1. Aberrations in these pathways create a physiology of GHI. The GHR is composed of three domains: an extracellular, hormone-binding domain; a single membrane-spanning domain; and a cytoplasmic domain. Some metabolic actions of GH, as well as its clearance, appear to be closely associated with GHBP concentrations. However, whereas GHBP levels generally reflect GHR levels and activity, levels of GHBP correlate only modestly with 24-hour GH production and growth rate. It is postulated that this reflects adjustments of GH secretion to accommodate GHR levels that may be genetically determined or modulated by environmental factors such as nutritional status. Levels of GHBP are low in early life, rise through childhood, and plateau during the pubertal years and adulthood. Impaired nutrition DM, hypothyroidism, chronic liver disease, and a spectrum of inherited abnormalities of the GHR are associated with low levels of GHBP, while obesity, refeeding, early pregnancy, and estrogen treatment can cause elevated levels of GHBP.

   Patients with GHI have the phenotype of GHD and diminished production of IGF-1, but with normal or elevated serum GH levels (Figure 2-25).¹⁰ Measurement of abnormally low serum levels of GHBP is useful in identifying some subjects with GHI caused by genetic abnormalities of the GHR. However, patients with GHI caused by nonreceptor abnormalities, defects of the intracellular domain of the GHR, or inability of the receptor to dimerize may, however, have normal or even high serum levels of GHBP. Primary GHI can occur from (1) abnormalities of the extracellular, dimerization, or intracellular domains of the GHR; (2) postreceptor abnormalities of GH signal transduction; (3) primary defects of IGF-1 biosynthesis; and (4) genetic insensitivity to IGF-1 action. Secondary GHI is an acquired and relatively common condition that can be caused by (1) malnutrition; (2) hepatic, renal, and other chronic diseases; (3) circulating antibodies to GH; and (4) antibodies to the GHR.

   Individuals with GHI show diminished response to exogenous hGH, in terms of growth, metabolic changes, or of increases in serum levels of IGF-1 and IGFBP-3.³¹ GHBP activity is undetectable in 75% to 80% of patients with this disorder. A wide variety of homozygous point mutations in this gene (missense, nonsense, and abnormal splicing) have been identified, with most existing in the extracellular or dimerization domain of the GHR. It is unclear whether substitutions identified in the intracellular GHR domain in short children represent genuine mutations or innocent polymorphisms.

   Heterozygosity for defects of the GHR may clinically suggest partial GHI, raising the question of whether some children labeled as “idiopathic short stature” may harbor such mutations. Given the requirement for dimerization of the GHR, an abnormal protein could exert varying degrees of dominant negative effect. However, inconsistencies in studies of results relating identified GHR mutations with low GH/IGF-1 axis effect and vice versa justify maintaining healthy skepticism regarding the ability to precisely dissect out patients with varying degrees of GHI from the ISS group.

   Replacement therapy for severe primary IGF-1 deficiency was approved by the FDA to be used only in the following cases: (1) mutations in the GHR gene, (2) post-GHR defects (including mutations in the STAT5b gene, which are also associated with immunodeficiency), (3) IGF-1 gene mutations, and (4) patients with GH-1 gene deletions who develop functional GH-inhibiting antibodies.⁴⁵ Approval of rhIGF-1 treatment was based on five clinical trials of which four were open-label studies and one was double-blind and placebo-controlled. These studies involved a total of 71 children and showed that administration of rhIGF-1 increased height velocity compared to baseline in the first year by 5.2 cm/year, second year 2.9 cm/year, third year 2.3 cm/year, and fourth to sixth years 1.5 cm/year. Treatment also increased bone maturation rate, with BA increasing 8.1% faster than chronological age. The FDA has specified that rhIGF-1 may be used to treat severe primary IGF-1 deficiency cases that meet the following criteria:

   • Height SD is ≤ 3.0 for the child’s age and sex.

   • Basal IGF-1 SD is ≤ 3.0.

   • Normal or elevated GH, except for children with GH gene deletion.

   • All indications of secondary IGF-1 deficiency have been ruled out such as GHD, malnutrition, chronic treatment with anti-inflammatory steroids, and hypothyroidism.

   • Contraindications for treatment include age < 2 years, closed epiphyses; active or suspected neoplasia; chronic illness such as DM, cystic fibrosis (CF), etc; or allergy to IGF-1 or any of the other ingredients used to stabilize the medication.

   Dosage should be individualized for each patient. The recommended starting dose of rIGF-1 is 0.04 to 0.08 mg/kg (40-80 μg/kg) twice daily by subcutaneous injection. If well-tolerated for at least 1 week, the dose may be increased by 0.04 mg/kg per dose, to the maximum dose of 0.12 mg/kg given twice daily. Preprandial glucose monitoring is recommended at treatment initiation and until full dosing is established. If frequent symptoms of hypoglycemia occur, preprandial glucose monitoring should continue. Adverse events occur in approximately 60% of treated patients, including, most commonly, headache (38%), vomiting (25%), and hypoglycemia (14%). Adverse events are usually transient, rarely lead to treatment discontinuation, and are without known sequelae.⁴⁶

5. IGF-1 deficiency due to post-GHR abnormalities and IGF-1 insensitivity

   At multiple sites, post-GHR defects in translation of GH message can result in growth impairment.⁴⁷ Defects in JAK2 and MAPK leading to activate the STAT pathway have been described in children with profound short stature, failure to respond to hGH, and normal GH binding and GHBP levels. Primary defects in IGF-1 biosynthesis caused by IGF-1 gene deletions can lead to severe pre- and postnatal growth retardation, sensorineural deafness, insulin resistance (caused by high GH levels), insensitivity to exogenous hGH, but reversal (albeit often incomplete) of growth and metabolic abnormalities with exogenous IGF-1 treatment. Conditions of insensitivity to IGF-1 action also exist and include abnormalities of (1) IGF-1 transport and clearance that would alter presentation of IGF-1 to its receptor, (2) the IGF1R itself, and (3) postreceptor signaling activation. Such patients would be expected to be exceedingly small, both during pre- and postnatal life, have elevated GH and normal to high IGF-1 levels, and poor growth responses to hGH and (presumably) IGF-1 administration. Primary defects of IGF transport and clearance, apparent deficient IGF-1 receptor production (eg, African Efe pygmies), or responsiveness (eg, leprechaunism) have also been reported as rare causes of growth failure.

Diagnosis.

The foundation for the diagnosis of IGF-1 deficiency is careful documentation of serial heights and determination of height velocity. Even in children below the 5th percentile in height, documentation of a consistently normal height velocity (above the 25th percentile) makes the diagnosis of either IGF-1 deficiency or GHD highly unlikely. Key historical and physical examination findings indicating that GHD could be present are listed in Table 2-6.

1. Assessment of GH secretion

   The assessment of pituitary GH production is difficult because its secretion is pulsatile with the most consistent surges occurring at times of slow-wave EEG rhythms during phases 3 and 4 of (deep) sleep (see earlier in this chapter). Between normal pulses of GH secretion, serum GH levels are low (often < 0.1 μg/L), below the limits of sensitivity of most conventional assays (usually < 0.2 μg/L). Accordingly, measurement of a random serum GH concentration is virtually useless (beyond the immediate newborn period) in diagnosing GHD but may be useful in the diagnosis of GHI and GH excess. Instead, physiologic and pharmacologic stimuli are used to assess GH secretion ability. Physiologic stimuli include fasting, sleep, and exercise. Pharmacologic stimuli include clonidine, glucagon, propranolol (all of which act by increasing GHRH), as well as arginine and insulin (which suppress somatostatin). Stimulation tests can be divided into “screening tests” (exercise, fasting, and clonidine), which are characterized by ease of administration, low toxicity, low risk, and low specificity, and more “definitive tests” (insulin, glucagon, and arginine). To improve specificity, provocative tests are performed in a fasting state (which reduces IGF-1 levels, increases ghrelin, and minimizes GH secretory suppression by glucose), and sometimes sequentially (though this has not been shown to improve specificity). It is commonly required that a child “fail” two distinct provocative tests to be diagnosed with GHD. Standard provocative GH tests are summarized in Table 2-7.

   Provocative GH testing, however, has a number of pitfalls. None of the standard pharmacologic provocative tests satisfactorily mimics the normal secretory pattern of pituitary GH. Even when naturally occurring regulatory peptides are used for stimulation, their dosage, route of administration, and interactions with other regulatory factors are artificial. Second, the definition of a “subnormal” response to provocative tests is arbitrarily defined, and has varied with time and by site. Before recombinant hGH was available, a cutoff level of 2.5 ng/mL had often been used. In the era of plentiful therapeutic GH, this cutoff was increased to 7 ng/mL and then to (the currently used) 10 ng/mL, although there are no data for validating higher arbitrary cutoff values. Third, marked variability of GH assays (as much as threefold variability in the measurement of serum GH levels among established laboratories) limit their discriminatory power. Further, newer GH assays measure GH immunopotency at 30% to 50% of earlier assays, but “new normal” cutoff levels have not been defined. Fourth, GH secretion normally varies with age, and prior to puberty and during the early phases of puberty, GH secretion may normally be so low as to blur the discrimination between GHD and CDGP. Many children who “fail” provocative testing before the onset of puberty prove to have “normal” GH secretion after puberty or after administration of exogenous sex steroids.⁴³ Though it is well documented that such “priming” with sex steroids helps identify the likely return of normal GH secretion with puberty, the role of such manipulations in guiding management of the poorly growing preadolescent remains obscure and is vigorously debated. What is clear is that reliance on provocative testing alone without consideration of this peripubertal phenomenon can lead to erroneous and excessive diagnosis of GHD. Finally, provocative testing typically requires multiple timed blood samples and often the parenteral administration of drugs. The resulting discomfort, occasional risk to the patient, and expense are self-evident.

   Alternative approaches involve analysis of spontaneous GH secretion by multiple sampling (every 5-30 minutes) or by continuous blood withdrawal over 12 to 24 hours. The former method allows one to evaluate and characterize GH pulsatility, whereas the latter only permits determination of mean GH concentration. In spite of these apparent advantages, spontaneous GH secretion identifies only about one-half of the children with GHD as defined by provocative testing and approximately 25% of normally growing children show low overnight GH levels. These issues, combined with the obvious expense and discomfort of such testing, have precluded its widespread acceptance as a practical diagnostic alternative.

2. Measurement of the GH-dependent peptides

   An important alternative means of diagnosing GHD is the assessment of IGFs and their binding proteins. IGF-1 levels are more GH-dependent than are IGF-2 levels and are more likely to reflect subtle differences in GH secretory patterns. However, serum IGF-1 levels are influenced by age, degree of sexual maturation, and nutritional status. IGF-1 levels in normal children younger than 5 years are low with overlap between the normal range and values in GH-deficient children. Assessment of serum IGF-2 levels is less age-dependent, especially after 1 year of age, but IGF-2 is less GH-dependent than is IGF-1. Serum levels of both peptides are relatively constant during the day and are not affected by acute fasting, so that no special preparation is required for the blood draw.

   Limitations of reliance on IGF-1 measurement include (1) interfering substances in radioimmunoassays, radioreceptor assays, and bioassays by IGFBPs, which must be removed or blocked; (2) age dependency, being lowest in young children (< 5 years of age), a period during which an accurate and easily obtained diagnostic test is most wanted; (3) levels may be low in conditions other than GHD, such as primary GHR dysfunction (Laron syndrome) and secondary GHI (malnutrition, liver disease, etc); (4) levels of IGF-1 (and IGFBP-3) are frequently normal in adult-onset GHD and in children with GHD resulting from brain tumors and/or cranial irradiation; and (5) there is imperfect correlation between serum IGF-1 levels and provocative or spontaneous GH measurements. In short-statured children younger than 10 years, IGF-1 levels are below −2 SD in only approximately 50%. In fact, IGF-1 levels alone allow discrimination between GHD and normal short children only in children with BAs of more than 12 years (ie, postpuberty onset). When serum levels of both IGF-1 and -2 are determined, the correlation with GH testing improves, because serum IGF-2 levels are low in GHD and normally do not increase with age after 1 year. The observation that “normal short” children may have low serum levels of IGF-1, IGF-2, or both calls into question the criteria by which the diagnosis of GHD is made.

   Given that provocative GH testing is both arbitrary and nonphysiologic and the inherent variability in GH assays, it is not surprising that the correlation between IGF-1 levels and provocative GH levels is imperfect. These points are further supported by recent observations with immunoassays for IGFBP-3. Measurement of IGFBP-3, normally the major serum carrier of IGF peptides, has clinical value because it is GH dependent and its concentrations correlate with the sum of the levels of IGF-1 and -2. Serum IGFBP-3 levels vary with age to a lesser degree than is the case for IGF-1. Even in infants, serum IGFBP-3 levels are sufficiently high to allow discrimination of low values from the lower end of the normal range. Serum IGFBP-3 levels are also less dependent on nutrition than is IGF-1. Most important, IGFBP-3 levels are GH dependent, with 97% of children diagnosed with GHD by conventional criteria (height < 3rd percentile, height velocity < 10th percentile, and peak serum GH < 10 ng/mL) having IGFBP-3 levels below the 5th percentile for age whereas 95% of non-GHD short children had normal IGFBP-3 levels. However, such a clear correlation between provocative GH testing and serum IGFBP-3 levels is not observed in clinical practice. There is the additional concern that IGFBP-3 levels may be “falsely” normal in patients with GHD, resulting from acquired intracranial lesions.

   In summary, diagnosis of GHD is challenging and requires careful and collective analysis of multiple factors. In a child with attenuated growth and delayed skeletal maturation, a low IGF-1 level (for BA) accompanied by low provoked GH levels (eg, < 5 ng/mL) strongly suggests GHD; pituitary MRI can be helpful in such cases, because abnormal findings such as diminished pituitary size or ectopic posterior pituitary enhancement strengthen the diagnosis, whereas normal pituitary/hypothalamic anatomy weakens it.

3. Diagnosis of IGF deficiency caused by GHI

   The combination of decreased serum levels of IGF-1, IGF-2, and IGFBP-3, along with increased or normal serum levels of GH, suggests the diagnosis of GHI; namely, (1) basal serum GH greater than 5 ng/mL; (2) serum IGF-1 50 μg/L or less; (3) height SDS −3 or less; (4) serum GHBP less than 10% (based on binding of [¹²⁵I]GH); and (5) a rise in serum IGF-1 levels after GH administration of less than twofold the intra-assay variation (∼10%).⁴⁸ Unfortunately, in practice, there is striking discordance between serial studies in ISS patients for both IGF-1 and IGFBP-3 generation in response to hGH, affirming the difficulty of assigning firm cutoff values of normal GH sensitivity. Decreased serum levels of GHBP also suggest the diagnosis of GHI. As mentioned previously, though, individuals with GHI caused by mutations in the dimerization site or in the intracellular domain of the GHR or abnormalities of postreceptor signal transduction mechanisms have normal serum concentrations of GHBP. Consequently, definitive diagnosis of GHIrequires (1) the classic phenotype; (2) decreased serum levels of IGF-1 and IGFBP-3; and (3) identification of an abnormality of the GHR gene.

4. Possible partial GHI: idiopathic short stature and heterozygous defects of the GHR

   Many children and early adolescents are short (< 3rd percentile), have slowed linear growth velocity (< 25th percentile), may have delayed skeletal maturation and an impaired or attenuated pubertal growth spurt, with or without a family history manifesting some or all of these clinical features, and have no chronic illnesses or apparent endocrinopathies. Such children usually have normal GH secretory dynamics, though provocative tests may be blunted under some circumstances; GH-dependent peptides are lower than expected on a chronological basis (though usually not if corrected for skeletal age); and treatment with hGH usually augments linear growth. Such children are usually considered variants of normal growth and achieve a final adult height within the range considered acceptable for the family. The etiology of the slowed childhood growth and frequently delayed pubertal growth spurt has not been established in most of these children. Eventually, causes of ISS will likely be identified at each level of the hypothalamic-pituitary-GH-IGF-1-growth plate axis.

   Children with heterozygous mutations of the GHR have been considered as candidates to explain some cases of ISS. Serum levels of GHBP in most children with ISS are lower than the normal mean and are even lower in children with low IGF-1 and higher mean 12-hour levels of GH. These observations raise the possibility that an abnormality of GHR content or structure could impair GH action. In heterozygotes, protein from the mutant allele may disrupt the normal dimerization that occurs when GH interacts with its receptor, leading to diminished GH action and growth impairment. So far, however, analysis of GH-induced IGF-1/IGFBP-3 generation and genes in ISS children affirm that such heterozygotes explain only a small proportion of patients. Post-GHR and post-IGF1R dysfunction remains a potentially fertile area in which to uncover more causes of “idiopathic” growth failure.⁴⁷

Treatment.

In general, treatment of IGF-1 deficiency should begin with a trial of hGH therapy. Subcutaneous daily hGH administration is currently preferred, with most children with GHD in the United States receiving daily treatment with 0.04 to 0.05 mg/kg/day, although many children show early satisfactory growth on smaller doses. Current hGH preparations are essentially equivalent. Because growth before puberty is a major determinant of adult height, early initiation of GH treatment allows more complete normalization of height. Whereas there can be a temptation to defer injection therapy in young children in order to minimize discomfort and inconvenience, available evidence strongly supports early recognition, referral, diagnosis, and treatment of severely GH-deficient patients as an important step toward optimizing growth potential.

Increasing the dose of hGH improves growth rate; comparison of 0.025, 0.05, and 0.1 mg/kg daily in prepubertal severely GH-deficient children showed significantly greater growth velocities and gains in cumulative height SDS in the 0.05 and 0.1 mg/kg/day groups compared with the 0.025 mg/kg/day group after 2 years of treatment. There were no significant differences between the 0.05 and 0.1 mg/kg/day groups.⁴⁹ Increased frequency of hGH administration improves growth rate, suggesting the “pulsing” message of GH to its target cells, in addition to adequacy of GH levels, and enhances linear growth. Nocturnal administration, which more closely mimics physiologic GH secretion, could theoretically add to efficacy, but this is not consistently observed.

The serum IGF-1 response to hGH is dose dependent, continuing to rise as hGH is administered in doses above currently prescribed levels. Epidemiological studies that suggest an association between high serum IGF-1 levels and the incidence of malignancy have prompted a recommendation that IGF-1 and IGFBP3 levels be monitored on a regular basis. A younger age, greater delay in height age and BA, and greater severity of GHD based on provocative testing each correlate with improved initial response to hGH therapy. In multiple regression analysis, variables that have positive effects on adult height of GH-treated patients include taller parents, more frequent hGH injections, longer duration of hGH treatment, taller height at start of hGH treatment, and greater severity of GHD.

Regardless of the regimen chosen, the effect of hGH wanes with time, and the first year of treatment usually produces the greatest growth increment. After this early phase of rapid growth, short-term increased replacement doses of hGH renew catch-up growth without adverse on-treatment metabolic effects. Seasonal variation in growth rate during hGH therapy, with peaks in the summer and nadirs in the winter (North American population), has also been described.

Children who have had removal of craniopharyngioma frequently experience growth acceleration in the absence of measurable GH. Patients with hypothalamic involvement generally achieve normal height more often than patients without hypothalamic involvement. This phenomenon, known as “growth without GH,” is still not fully understood, but may be attributable to postoperative hypothalamic dysfunction resulting in nutritional excess and hyperinsulinemia. Insulin has known anabolic effects and, at high concentrations, it may bind to the IGF1R to induce growth. However, children postcraniopharyngioma treatment who are growing despite GHD do not appear to differ significantly from those requiring hGH treatment in terms of anthropometrical measures, body composition, and metabolic indexes, including insulin levels. Finally, the effects of other local growth factors (eg, GH variants and IGFs) acting on bone metabolism have been hypothesized to contribute to the phenomenon of “growth without GH.”⁵⁰ Polyphagia and significant weight gain are usually also observed. Though supplementation with hGH is not required to sustain linear growth, body composition analysis reveals increased fat mass and decreased lean mass typical for the GH-deficient state. Unfortunately, postsurgical treatment with hGH, while promoting linear growth, does not appear to slow the excessive weight gain commonly seen in this population. Avoidance of excessive cortisol-replacement therapy is extremely important in these individuals. This growth pattern may persist, allowing attainment of normal adult stature without hGH therapy, but hGH may nevertheless be indicated to improve body composition and other metabolic consequences of GHD.

The issue of whether hGH replacement doses should be increased substantially during puberty remains debated. While clearly insufficient dose and frequency of hGH administration may reduce expected adult height, most studies show that the hGH dose administered during puberty to be a relatively minor factor in determining total pubertal growth as compared to gender, chronological age at pubertal onset, and prepubertal height relative to mid-parental height. An increase in dose to 0.7 mg/kg/week (in contrast to conventional dosage recommendations of 0.18-0.35-mg/kg/week) has been shown to improve growth rates, near-adult height, and height SDS in adolescents with GHD without evident on-treatment adverse effects.⁵¹ However, the cost of such treatment and the induction of IGF-1 levels in the supraphysiologic range and potential long-term consequences must be balanced with the possible added benefit achieved.

With early diagnosis, careful attention to accompanying hormonal deficiencies, and progressive dose adjustments, children with GHD usually reach normal adult height (Figure 2-26). The BA will advance with hGH treatment, but usually not more than height age. Linear growth often accelerates faster than BA following initiation of hGH therapy, leading to increases in predicted adult height. Even with successful long-term hGH therapy, however, correction of disabling short stature does not consistently normalize the psychosocial outcome for adults with GHD.

Failure to demonstrate an increase in growth rate and IGF-1 levels in response to confirmed consistent administration of hGH therapy may be secondary to nonadherence; however, the diagnosis of GHI and rationale for additional diagnostic tests (eg, GHBP and/or anti-GH antibody analysis) should be considered. Indications to the use of rhIGF-1 as a “second-tier” therapy include GHI (Laron syndrome), in which severe short stature and other characteristic features (see Figure 2-25) are the result of GHI caused by GHR dysfunction as mentioned above.

The consequences of severe GHD in adult life and the beneficial effect of replacement therapy are increasingly well established. Accurate selection of appropriate candidates for adult hGH treatment and the transition of their care from pediatrics to adult medicine require careful consideration of several issues. Because the majority of children who are diagnosed with GHD and treated with hGH do not have permanent GHD, anticipatory counseling regarding possible lifelong treatment should be focused on children with panhypopituitarism and those with severe IGHD associated with CNS abnormalities. Appropriate timing for termination of “growth-promoting” hGH therapy should be guided by efforts to balance the high cost of late-adolescent treatment with the attainment of reasonable stature goals.

Confirmation of GHD following cessation of hGH therapy for at least 1 month and provocation with appropriate stimuli (ie, insulin, arginine, or glucagon) and measurement of IGF-1 levels are appropriate for all candidates for adult hGH therapy who do not have well-documented multiple pituitary hormone deficiencies. Testing of the GH axis can be performed within weeks of hGH cessation, but confirmation of an emerging adult state of GHD with body composition, lipid panel, bone densitometry testing, and quality-of-life assessments may require 1 or more years of posttreatment observation. Though it remains unclear whether such a period of observation is necessary or advisable for late adolescents with suspected panhypopituitarism, a 6- to 12-month trial off hGH does not appear to be detrimental. Those adolescents with a history of less severe GHD or those with “partial GHD” (ie, stimulated GH values of 5-10 ng/mLon retesting as an adult) pose a greater diagnostic and therapeutic challenge. If therapy is interrupted in this group, a minimum of close clinical, radiologic (if not previously performed), and laboratory monitoring for evidence of an evolving state of GHD and/or other pituitary hormone deficiencies is required.

Hypothyroidism

Etiologies and effects on growth.

Hypothyroidism (see Chapter 4) can be congenital or acquired. The congenital form (CH) is most commonly associated with aberrant embryologic glandular migration, resulting in an absent or poorly functioning ectopic gland. Less commonly (∼10% of cases), inherited enzyme deficiencies interfere with thyroid hormone synthesis. Untreated severe CH leads to jaundice, growth failure, anemia, and neurologic damage. Fortunately, newborn screening programs have nearly eliminated the morbidity of severe undiagnosed CH in developed countries. Less commonly, CH results from deficient pituitary or hypothalamic function, usually more mild in severity and accompanied by other hormonal deficiencies. Acquired hypothyroidism is most commonly caused by chronic lymphocytic thyroiditis (Hashimoto thyroiditis); other causes include radiation exposure, iatrogenic surgery, or medications (eg, lithium).

During early gestation, the developing child is dependent on maternal thyroid hormone in utero for normal neurologic development, but less so for normal growth. After birth, in contrast, linear growth is heavily dependent on normal serum levels of T₄, and severe untreated CH leads to profound growth failure. In acquired states of thyroid deficiency, which usually evolve gradually, growth retardation may be subtle at first, but it usually progresses in severity over ensuing years; that is, hypothyroidism should always be strongly considered in the child demonstrating near-complete cessation of growth. Proportionally, linear height is more affected than is weight acquisition, creating a growth-stunted and overweight-for-height child (Figure 2-27). Skeletal maturation is invariably delayed commensurate with the duration of thyroid hormone deficiency, and body proportion is immature with increased upper-to-lower segment ratios.

Diagnosis and treatment of hypothyroidism is described in Chapter 4. Importantly, growth failure should not be attributed to “compensated hypothyroidism” when laboratory studies reveal normal free thyroxine (T₄) and triiodothyronine (T₃) levels with a mildly elevated TSH. Treatment leads to prompt resumption of normal or supranormal growth rates. Unfortunately, replacement therapy may not fully restore growth potential (especially if treatment is initiated near puberty) because of the associated rapid skeletal maturation that ensues once treatment is initiated. The deficit in adult height correlates with the duration of hypothyroidism. In severe cases, pharmacologic delay of puberty and growth plate fusion to improve height prognosis may be a consideration.

Glucocorticoid excess

Glucocorticoids interfere with normal bone growth and metabolism at multiple steps in the growth cascade: augmentation of somatostatin tone and suppression of GH secretion, impairment of IGF-1 bioactivity, suppression of new collagen synthesis and osteoblastic function, and enhancement of bone resorption and protein catabolism (detailed in the section Chronic Disease and Glucocorticoid Therapy, later).⁵²

Exposure to glucocorticoids in excess of normal physiologic production can occur (rarely) from endogenous production (eg, pituitary adenoma, adrenal tumor) or exogenous therapeutic preparations (eg, used for the treatment of asthma, juvenile rheumatoid arthritis, renal transplantation, or inflammatory bowel disease [IBD]—see discussion on chronic illness as follows) (Figure 2-28). Cushing syndrome in childhood (described in detail in Chapter 5), whether caused by an ACTH-secreting tumor or an adrenal neoplasm, usually presents with growth failure accompanied by excessive weight gain (Figure 2-29). This effect on linear growth is a key finding differentiating glucocorticoid excess from the normal or accelerated linear growth resulting from exogenous obesity. Children with adrenal tumors may produce excess androgens as well as glucocorticoids, so that growth deceleration may be less apparent owing to the concomitant growth-promoting effects of androgens.

Calcium, phosphate, and vitamin D abnormalities

The regulation of calcium, phosphate, and vitamin D levels and interactions is critical for formation and growth of normal bones. In a growing child, disruptions at any point in the parathyroid-vitamin D-kidney-bone system can result in skeletal abnormalities and short stature. Examples include dietary deficiencies or excessive losses of vitamin D, parathyroid hormone (PTH) resistance syndromes, and disorders resulting in low levels of phosphate such as X-linked hypophosphatemia (XLH) (see Chapter 7 for details). Inherited disorders are often suggested by skeletal abnormalities and disproportional features in the parents.

Parathyroid resistance syndromes include a heterogeneous group of disorders that involve a state of hypocalcemia and hyperphosphatemia secondary to end-organ resistance to PTH. The genetic defect (when detectable) is in the GNAS1 gene encoding the α-subunit of the stimulatory G-protein leading to a diminished/absent response to PTH. Depending on the specific postzygotic mutation, variable degrees of resistance that may be present in various hormonal axes utilizing the G-protein signaling pathway (TSH, ADH, gonadotropins, glucagon, ACTH, and GHRH) may also be affected. Children with specific forms of PTH resistance (types 1a and 1c, pseudo-pseudohypoparathyroidism) usually have short stature owing to intrinsic bone abnormalities, but moderate degrees of accompanying hypothyroidism and/or GHD should be excluded as potential contributors to growth failure. Distinctive shortening of metacarpals may be present, along with truncal obesity and mental retardation. Collectively these features are referred to as Albright hereditary osteodystrophy (AHO). Treatment of parathyroid resistance includes calcium supplementation, phosphate restriction, and activated vitamin D in the form of calcitriol.

In vitamin D–insufficient states, there is a reduction in the intestinal reabsorption of calcium resulting in low serum ionized calcium levels and a triggering of the release of PTH. Increases in PTH levels result in (1) increase in calcium reabsorption in the kidneys; (2) activation of 1-α-hydroxylase, which increases the conversion of vitamin D to its active form of 1,25-dihydroxyvitamin D; and (3) the release of calcium from bone. Elevated PTH levels stimulate phosphaturia that, along with reduced phosphate absorption without active vitamin D, causes hypophosphatemia and low phosphate in bone which then leads to failure of mineralization referred to as rickets in growing bone and osteomalacia in mature bones. In spite of food supplementation, vitamin D insufficiency continues to be a cause of rickets and growth retardation, particularly in infants born prematurely or to vitamin D–deficient mothers, especially if breast-fed exclusively for several months and exposed to low natural light. Characteristic findings include frontal bossing, craniotabes, rachitic rosary, bowing of the legs, and gross motor delays.

Phosphorous is also an important component in the physiologic processes involved in bone mineralization. A hypophosphatemic state leads to decreased mineralization and disorganization of the growth plates resulting in skeletal abnormalities. Various causes of low serum phosphorous include vitamin D deficiency (see the preceding discussion) and inherited disorders that result in the increased renal loss of phosphate. One such disease is X-linked hypophosphatemic rickets (XLHR), a dominant disorder caused by inactivating mutations in phosphate-regulating gene with homologies to endopeptidases on the X chromosome (PHEX) that is associated with increased circulating FGF-23 levels. Ultimately children with XLH have abnormal renal phosphate reabsorption and disruption of vitamin D metabolism (ie, lower than expected 1,25-dihydroxyvitamin D levels with normal PTH in the setting of hypophosphatemia), resulting in disproportionately short stature and radiologic evidence of rachitic disease. Serum phosphate levels are low, but calcium levels are normal. Conventional treatment has included oral inorganic phosphate salts and activated vitamin D (usually calcitriol). Treatment during growth partially corrects leg deformities, decreases the number of surgeries, and may improve adult height. Early initiation of treatment appears to optimize height outcomes. Although an impaired GH-IGF-1 axis is not a primary cause of short stature in XLHR patients, the physiologic renal antiphosphaturic effect of GH may be an adjunctive tool, associated with the conventional treatment, to improve linear growth in poorly growing XLHR patients, without worsening of the degree of body disproportion. However, the results are not conclusive, as some patients do not have catch-up growth during hGH treatment.⁵³

Abnormal Growth in Chronic Disease States

Growth impairment frequently accompanies chronic illness in children and adolescents (Table 2-8). In most circumstances, growth impairment is the result of multiple abnormalities. In the initial response to an acute illness, the anterior pituitary actively continues to secrete hormones into the circulation while, in the periphery, anabolic target organ hormones are inactivated. This response is thought to be beneficial and adaptive, causing reduced energy consumption; redirecting substrate use to make glucose, amino acids, and FFAs preferentially available to critical organs; postponing catabolism; and inducing activation of the immune response. However, when critical illness becomes prolonged, this initial positive adaptation becomes progressively more deleterious. For example, during chronic illness pulsatile secretion of anterior pituitary hormones becomes uniformly reduced secondary to reduced hypothalamic stimulation, leading to decreased protein synthesis and increased protein degradation, reduced fat oxidation, insulin resistance resulting in impaired anabolism, and reduced activity of target organs. In most cases, this chronic maladaptation is combined with factors related directly to the primary illness itself. In rat models for IBD, for example, about 60% of the final growth impairment is attributable to undernutrition, whereas inflammation accounts for the remaining growth deficit. In humans, glucocorticoid treatment often further compromises growth.

Causes and effects of imbalance betweenenergy supply and needs

Normal growth requires a balance between nutrient requirement and intake. Patients with CF, for example, either lose weight or fail to grow normally if their absorbed energy intake is less than their total daily energy expenditure.

Reduced intake in chronic diseases.

Reduced macro- and micronutrient intake and a disordered metabolic state result in both an insufficiency of substrate and abnormalities in enzyme function. A number of specific factors may contribute to reduced energy intake, including anorexia, mechanical problems, pain, malabsorption, and increased losses.

Anorexia.

The melanocortin system in the hypothalamus and brain stem has a central role in the regulation of feeding behavior. It is composed of two types of neurons situated in the arcuate nucleus of the hypothalamus: those that produce pro-opiomelanocortin (POMC) and induce an anorexic signal and those that produce the agouti-related peptide (AgRP) and neuropeptide Y (NPY) and induce an orexigenic signal. POMC is cleaved into α-melanocyte–stimulating hormone (α-MSH), which acts on the melanocortin-3 receptor (MC3R) and/or the melanocortin-4 receptor (MC4R). The AgRP/NPY neurons release AgRP, which is a natural antagonist of the MC3R and MC4R, and blocks the action of α-MSH. Together, the dynamic interaction between POMC and AgRP synthesis and secretion promotes feeding stimulation or inhibition. Inflammatory cytokines can influence the brain by modulating peripheral neurons that project to the brain through the vagus nerve, controlling secretion of adipocyte hormones such as leptin, and possibly acting directly in the brain, through the local production of cytokines and chemokines. The increased levels of cytokines stimulate the central melanocortin system and induce loss of appetite. Melanocortin inhibition has been shown to be a powerful tool in blocking the symptoms of cachexia.

Mechanical problems and pain.

Children with genetic syndromes frequently have feeding problems and swallowing dysfunction as a result of a complex interaction of anatomical, physiologic, and behavioral factors. Oral dysfunction (cleft lip/palate), neuromotor impairments such as in hypotonia, and pain may lead to feeding difficulties.

Malabsorption.

Growth retardation may reflect chronic malabsorption further complicated by increased catabolism. For example, failure of normal growth in a child with IBD is an indicator of insufficient and unsuccessful anti-IBD therapy. Growth resumes after effective control of the disease, reduction in inflammation, and improvement in nutritional intake.

Increased loss.

Chronic diarrhea and vomiting, such as that occurring in chronic regurgitation and gastroesophageal reflux in infancy, cause poor weight gain and growth. In diabetes, poor glycemic control and a catabolic tendency secondary to underinsulinization may also result in poor growth. Similarly, impaired insulin action/response and chronic glucosuria may compromise nutrition status and worsen growth in patients with CF-related diabetes.

Increased energy needs in chronic diseases.

Caloric expenditure is often increased in children with chronic disorders owing to increased cardiac or pulmonary work in congenital heart or lung disease, chronic infections in CF, chronic inflammation in Crohn disease (CD) or juvenile idiopathic arthritis (JIA), or increased protein catabolism due to glucocorticoid therapy. Lung inflammation, in particular, has been associated with increases in resting metabolic rate, and a human immunodeficiency virus (HIV)–associated hypermetabolic state has been described in adults and children.

Growth-disrupting endocrine effects of malnutrition and chronic illness.

Undernutrition leads to a fundamental shift in endocrine homeostasis, geared toward conserving energy and promoting diversion of substrate away from growth and reproduction toward critical body processes.

GH-IGF-1 axis.

In patients suffering from chronic undernutrition, both relative GH resistance with low IGF-1 and IGFBP-3, and reduced GH secretion have been reported. In addition, energy deficiency can inhibit the hypothalamic-pituitary-reproductive axis, leading to delay in pubertal development or frank hypogonadotropic hypogonadism, furthering reducing adult height due to blunting of pubertal growth.

• GH resistance. Stunting of body growth is a significant complication of advanced chronic renal failure (CRF). However, circulating GH levels are normal-to-elevated in uremia, suggesting the presence of resistance to GH action. Indeed, this disease is the best-studied example of GH resistance and serves as a good model for several other chronic disorders. Insensitivity to GH is the consequence of multiple defects in the GH/IGF-1 system.⁵⁴

• Loss of GH pulsatility. In patients who are critically ill for a prolonged time, the pulsatile release of GH is suppressed, whereas the nonpulsatile fraction of GH release remains somewhat elevated. Loss of GH pulsatility contributes to the low serum concentrations of IGF-1, IGFBP-3, and ALS. The origin of the loss of GH pulsatility resides within the hypothalamus and appears to be related to deficiency or inactivity of endogenous GHS, rather than to GHRH deficiency or resistance.

• Changes in GH receptor expression. Serum GHBP, which is a cleaved product of the GH receptor (GHR), is a marker for GHR density in tissues. GHBP is low in children and adults with CRF and proportionate to the degree of renal dysfunction and to growth rate.

• Janus kinase/signal transducer and activator of transcription (STAT) signaling. The binding of GH to the GHR results in receptor dimerization and autophosphorylation of the receptor tyrosine kinases and Janus kinase 2 (JAK2), which, in turn, stimulate phosphorylation of the STAT proteins (STAT1, STAT3, and STAT5). Upon activation, these STAT proteins translocate to the nucleus and activate GH-regulated genes. In uremia, a defect in postreceptor JAK2/STAT transduction has been documented. The JAK2/STAT pathway is regulated by, among other factors, suppressor of cytokine signaling (SOCS) proteins, which are induced by GH. These proteins bind to JAK2 and inhibit STAT phosphorylation. Upregulation of SOCS has been described in inflammatory states, and it may play a similar role in CRF.

• IGFs and IGFBPs. IGFs are transported in plasma bound to IGFBPs, which are responsible for extending their half-life and controlling their bioavailability. Of the six IGFBPs, IGFBP-1, IGFBP-2, and IGFBP-6, which are all inhibitors of IGF-dependent proliferation of chondrocytes, are elevated in CRF and correlate with the degree of renal dysfunction. Increased levels of IGFBP-1 and IGFBP-2 have been shown to correlate negatively with height. In addition to GH resistance, patients with CRF have poor growth during puberty as a result of delayed onset of pubertal growth, short duration of pubertal growth, and reduced gain of height during puberty. Reduced gain of height during puberty has been attributed to less frequent or absent LH pulses and reduced bioactivity of LH. These abnormalities are partially restored after kidney transplantation.

• GH deficiency. Deficiencies in anterior pituitary hormone secretion, ranging from subtle to complete, occur following radiation damage to the hypothalamic-pituitary axis (HPA), correlated with the severity and frequency of total radiation dose delivered and the length of follow-up. The GH axis is the most vulnerable to radiation.

• Ghrelin. The novel hormone, ghrelin—produced predominantly in the stomach and hypothalamus—is an endogenous agonist at the GHS receptor. In addition to its ability to stimulate GH secretion and gastric motility, ghrelin has a potent orexigenic effect mediated through hypothalamic NPY and AgRP, and increases the respiratory quotient (RQ). Therefore, defective ghrelin signaling from the stomach or hypothalamus could contribute to abnormalities in energy balance, growth, and associatedgastrointestinal and neuroendocrine functions. However, paradoxically, in patients with chronic illness, such as IBD, plasma concentrations of ghrelin are increased, correlate with the severity of disease, and show a negative correlation with IGF-1. These findings suggest that ghrelin resistance rather than deficiency may be characteristic of adaptation to chronic illness.

Leptin.

Leptin is a circulating hormone derived from adipose tissue that regulates energy intake, energy expenditure, and energy partitioning. In malnourished children, circulating leptin concentrations are low and correlate with low IGF-1 concentrations. In addition, low leptin concentrations suppress gonadotropin secretion and may increase cortisol concentrations.

Inflammation.

The release of inflammatory cytokines and, particularly in conditions characterized by wasting, leads to decreased protein synthesis in muscle and bone, development of GH and/or IGF-1 resistance, and enhanced protein degradation. Chronic inflammation, such as that occurring in patients with JIA and IBD, is characterized by the production of cytokines, such as tumor necrosis factor-α (TNF-α) and interleukin (IL)-1 and IL-6. In these disorders, growth retardation is strongly correlated with the severity and duration of the disease and serum concentrations of cytokines. High circulating IL-6 levels in transgenic mice overexpressing the IL6 gene induced a decrease in IGF-1 production and growth retardation. In addition, in patients with CD, either intestinal resection or enteral nutrition can induce catch-up growth in the early stages of puberty owing to remission of the inflammatory process and an early increase in IGF-1 levels preceding improvement in nutritional status.⁴⁷

Chronic hypoxia.

Chronic hypoxia is frequently associated with poor growth. Serum IGF-1 levels are lower and GH levels higher in patients with congenital heart disease than in normal controls, suggesting GH resistance. Moreover, serum IGF-1 levels are significantly lower in cyanotic CHD patients than in those who are acyanotic. Many oxygen-sensitive regulatory mechanisms work through hypoxia inducible factor-1 that plays a pivotal role in the adaptation to chronic hypoxia by allowing energy production in anaerobic conditions and that is also involved in the development of anorexia through induction of the promoter of the leptin gene.

Abnormal Growth in Specific Disorders

In the past, abnormal growth and delayed puberty were a frequent consequence of chronic diseases. However, with modern treatment, severe short stature is no longer common. However, in some specific situations, growth retardation remains a problem that is difficult to prevent and treat. In these cases, short adult stature will be a complication of the disease and will have an influence on the quality of life of the patient.

Diabetes

Clinical presentation.

Extreme short stature was a component of diabetes in the past and part of the “historical” Mauriac syndrome. For example, in a study of identical twins discordant for diabetes, it was found that the diabetic twin was, on average, 5.8 cm shorter than the control twin. Although modern management of diabetes has eliminated severe growth alterations, some degree of growth impairment can be observed in poorly controlled diabetic patients. Furthermore, after a period of better metabolic control, an improvement in the growth velocity has been described, an observation that can be used as an incentive for a better glycemic control by the young patient. Longer disease duration and poorer metabolic control are factors associated with a shorter adult height. This loss in height is, to a large extent, caused by a decrease in the pubertal growth spurt, particularly in girls.

Elevated plasma concentration of GH and decreased concentration of IGF-1 have been shown in patients with diabetes, and this apparent GH resistance may explain, in part, the abnormal growth observed in some patients; improved metabolic control has been associated with an increase in IGF-1. However, one should keep in mind that autoimmune disorders frequently associated with type 1 diabetes, such as hypothyroidism and celiac disease, may be present and be a major cause of growth retardation. These disorders should be excluded in any diabetic patients with abnormal growth velocity.

Treatment.

There is no specific treatment of short stature in children with diabetes. Improved compliance and metabolic control will result in increased growth rate and catch-up growth.

Chronic renal insufficiency

Immense progress has been made in the treatment of chronic renal insufficiency (CRI), but growth retardation remains a difficult problem in 30% to 60% of patients.

Clinical presentation.

There is no single cause of growth failure in CRI, and the underlying renal disease is important to consider when analyzing the mechanism of growth retardation. Congenital renal dysplasia is one of the most frequent causes of CRI during infancy and childhood. Children usually show a constant decline in renal function over time, and renal insufficiency is frequently associated with electrolyte and water loss, which contribute to growth failure. Malnutrition due to inadequate caloric intake and frequent vomiting may further exacerbate the poor growth. Children withglomerulopathies may also have growth retardation. The nephrotic state and glucocorticoid therapy in glucocorticoid-dependent nephrotic syndrome are both known causes of growth retardation in these patients.

In patients with tubular or interstitial nephropathies, the tubular dysfunction characterized by electrolyte, bicarbonate, and water loss can lead to severe growth retardation. The Fanconi syndrome is an example of complex tubular disorder that can be associated with severe stunting of growth. Cystinosis is another example of a disorder with severe short stature of complex etiology. Growth is compromised in infancy before alteration in glomerular function, and the decreased growth velocity observed at this age is mainly caused by tubular dysfunction. Generalized deposition of cystine crystals, particularly in the growth plate and in several endocrine glands, further aggravates the growth failure that is one of the most severe observed in children with CRF.

Although the uremic state is corrected by dialysis treatment, dialysis usually has a weak impact on growth. Renal transplant will improve the growth rate; however, the amplitude of the growth catch-up varies widely. Catch-up growth is influenced by the amount and duration of administered glucocorticoid, a strong argument for using alternate-day treatment, decreasing the overall dosage, and using alternative immunosuppressants to allow withdrawal of glucocorticoid therapy. Another factor is the age and glomerular function at transplantation. Indeed, several studies have shown a negative association between age at transplant and pubertal height gain. Because the total pubertal height gain is reduced in children with end-stage renal disease, early transplant will have a positive influence on pubertal height gain and adult height.

Treatment.

General measures, such as adequate caloric intake, alkali supplementation for the treatment of metabolic acidosis, and supplementation of water and electrolytes, have a strong influence on growth in infants and young children. Because the impact of dialysis and transplant on growth is far from optimal, the best strategy is to avoid the development of a growth deficit prior to transplant. This means early transplant, efficacious immunosuppressive therapy, and reduction (or avoidance) of glucocorticoid use. The efficacy of hGH treatment in improving growth velocity and adult height has been widely recognized, and this treatment has been approved in several countries. The effect of GH treatment on growth velocity is illustrated in Figure 2-30. The growth response is positively associated with residual renal function, duration of hGH treatment, and target height, and inversely associated with age at start of treatment. Long-term hGH administration results in increased adult height, but response is diminished in patients on dialysis and in those having delayed puberty.⁵⁵ Whether induction of puberty and intensified dialysis will improve outcomes needs more investigation. The safety of hGH therapy in this indication seems to be comparable to that shown for other indications.

Chronic inflammation

Clinical presentation.

Poor growth and short stature is a characteristic feature of chronic inflammation, particularly in patients with juvenile inflammatory arthritis (JIA). Although rare in the mono- or polyarticular forms, stunting of growth is part of the clinical picture in the severe systemic forms. As shown in Figure 2-31, decreased growth rate and short stature occur very early in the course of the disease and will lead to severe adult short stature. Chronic inflammation and glucocorticoid administration are the main causes of growth impairment.

Treatment.

Decreased glucocorticoid dosage and alternative therapy, such as anti–TNF-α, may positively affect the growth rate of these patients. Moreover, it has been shown that GH administration induces catch-up growth, improves final height, and, when given at onset of the disease, prevents the deleterious action of the disease on growth.^56,57 The safety of hGH in this setting seems to be excellent and, in the few studies reported so far, glucose tolerance seems only moderately affected. The positive effect of hGH on body composition, decreasing fat mass and increasing lean mass, is probably responsible for the modest anomalies of carbohydrate metabolism in these patients despite receiving glucocorticoids and hGH, the two drugs that induce insulin resistance.

Gastrointestinal diseases

Clinical presentation.

Several intestinal disorders may cause growth failure and decreased growth velocity may precede any specific signs of malabsorption or inflammation. Gastrointestinal disorders should therefore be considered in a child with unexplained growth failure. Celiac disease may lead to a failure of statural growth, although short stature is probably not a frequent sole manifestation in this disease. After institution of a gluten-free diet, the symptoms disappear rapidly and most patients exhibit complete catch-up growth.

Growth failure is one of the major complications in children with IBD. Although rare in children with ulcerative colitis, poor growth is a frequent manifestation in children with CD; short stature has been reported in 15% to 40% of adult patients with CD of pediatric onset. This large variation is probably explained by heterogeneity—extension of the lesions and degree of inflammation—of the cases included in the studies. The onset of puberty may be delayed and pubertal progression is frequently abnormally slow. Age at menarche is frequently late.

Treatment.

There is no specific treatment of the short stature caused by chronic inflammatory disease. Improving nutrition and decreasing the inflammatory state, however, will usually induce rapid catch-up growth. In CD, for example (Figure 2-32), it has been shown that constant enteral nutrition and partial intestinal resection will quickly decrease inflammatory indices, increase serum IGF-1, and ultimately increase growth velocity.

Chronic disease and glucocorticoid therapy

Glucocorticoids interfere with the GH-IGF-1 axis at the hypothalamic, pituitary, and target organ levels, affecting GH hormone release, receptor abundance, signal transduction, gene transcription, pre-mRNA splicing, and mRNA translation.⁵² In addition, glucocorticoids disturb normal calcium balance mediated through the intestine and kidney and have direct effects at the growth plate, including the suppression of gene expression; chondrocyte proliferation; and matrix proteoglycan synthesis, sulfation, release, and mineralization, as well as the augmentation of hypertrophic cell apoptosis. In addition, pharmacologic doses of glucocorticoids can induce a decrease in LH secretion.

The role of glucocorticoids in the treatment of certain forms of CRF and JIA has been described earlier, but there are several other situations where glucocorticoids are the key component of the treatment of a chronic disease and will be responsible for decreased growth and short stature. The severity of growth retardation mainly depends on the age at onset and severity of the disease and the duration and dosage of (ie, lifetimeexposure to) glucocorticoid therapy. In asthma, for example, the growth retardation is generally modest and the adult height is normal. This is mainly caused by improved use of inhaled glucocorticoids and decreased systemic exposure. On the other hand, in children with Diamond-Blackfan anemia, for example, short stature is present in one-third of the patients receiving prednisone, while being observed in only 5% of patients not receiving glucocorticoids.

Glucocorticoid administration not only has an impact on growth. Major changes in body composition such as muscle wasting, increased fat mass, and bone demineralization frequently develop and will have long-term deleterious effects in adults.

Treatment.

Decreasing glucocorticoid exposure, when possible, is the best way to diminish the consequences of treatment. The efficacy of hGH administration on linear growth has been reported in patients with glucocorticoid-induced growth retardation.⁵⁸ However, except for a few specific indications, evidence for a long-term beneficial effect of hGH on adult height is still lacking.

Thalassemia

Thalassemia major is an inherited hemoglobinopathy characterized by chronic anemia and iron overload due to transfusion therapy and gastrointestinal absorption. Iron overload and ectopic deposition—to which the anterior pituitary is susceptible—can disrupt multiple hormonal systems, resulting in hypogonadism, short stature, acquired hypothyroidism, and hypoparathyroidism. Disproportionate short stature is frequent and becomes more evident at puberty because of the lack of a growth spurt. Studies to date suggest that long-term treatment with hGH is ineffective in improving final height.⁵⁹

Metabolic Effects of GH in Children

Linear growth requires an anabolic state of protein synthesis and utilization of stored energy in fat. The enhancement of protein synthesis occurs, not surprisingly, in bone, cartilage, and skeletal muscle, but also in the erythropoietic system and other major organs (Figure 2-33). GH produces positive nitrogen balance, increases amino acid transport into cells, increases intracellular RNA, and decreases urea production and blood urea nitrogen (BUN) levels. A high-normal or mildly elevated BUN level and low serum phosphorus and alkaline phosphatase levels are usually observed in patients with GHD and normalize with hGH treatment. Markers of bone mineral metabolism are also improved. Children with GHD on hGH therapy show increased bone density and 1,25-dihydroxyvitamin D levels compared to pretreatment values.

GHRs and expression of GHR mRNA have been demonstrated both in preadipocyte cultures and in mature adipocytes. Some actions of GH in adipose tissue are mediated directly through interaction with the GHR, whereas others are mediated indirectly through IGF-1. These actions include (1) inhibition of differentiation of immature adipocytes to mature adipocytes; (2) enhancement of lipolysis and site-specific FFA release from adipose tissue; and (3) inhibition of lipoprotein lipase and stimulation of hormone-sensitive lipase. Consequently, GH-deficient children tend to demonstrate increased, predominantly abdominal, subcutaneous fat, which lessens and becomes more peripheral during therapy with hGH. Non–GH-deficient children also display reduction in overall body fat during hGH therapy.

The effects of GH on carbohydrate metabolism are complex and may be mediated indirectly via the antagonism of insulin action. GH is pivotal for maintenance of normal glucose homeostasis in infants, and GHD can cause life-threatening hypoglycemia beginning in the first week of life; interestingly, this critical role is markedly diminished in older children and adults. In a child with GHD, insulin secretion is diminished (related in part to pancreatic islet cell hypoplasia); however, insulin sensitivity is heightened. Nearly all studies of hGH therapy in children and adults show an increase in fasting and postprandial insulin levels. However, glucose homeostasis is generally not impaired. A theoretical concern is that normal levels of blood glucose and glycosylated hemoglobin during hGH therapy are preserved by compensatory hyperinsulinemia, which, when persistent, can be associated with atherosclerosis and hypertension. Higher levels of insulin (which do not correlate closely with the hGH dose) usually return to normal when therapy is discontinued. The association between hGH treatment and both types 1 and 2 diabetes is discussed as follows. Continued careful prospective follow-up of individuals during and after hGH therapy is needed to resolve this issue.

Potential Adverse Effects ofhGH Therapy

Recombinant hGH preparations are highly purified and free of contaminants. The possibility of infectious transmission through GH has been virtually eliminated. However, surveillance of patients who received pituitary-derived GH for development of CJD remains important. Antigenicity of hGH preparations is also low, although GH antibodies can be detected in 10% to 30% of treated children. With rare exceptions (< 0.1%), these antibodies do not impede the effects of hGH. Nevertheless, current dosages of hGH therapy commonly result in pharmacological rather than replacement therapy, prompting concerns about long-term effects of exposure to various degrees of GH excess during childhood.

Thyroid dysfunction

GH administration to healthy subjects acutely increases serum T₃ and reciprocally decreases free T₄ (FT₄). Similarly, transient laboratory indications of hypothyroidism may be seen in as many as 25% of GH-deficient children treated with hGH, with declines in FT₄ levels reflecting increased peripheral conversion of T₄ to T₃. Similarly, in patients receiving concomitant hGH and thyroid hormone–replacement therapy, attempts to achieve above-the-mean FT₄ levels can lead to elevated T₃ levels. Patients with GHD who display subnormal nocturnal TSH surges, signifying a preexisting central hypothyroidism, are more likely to display subnormal total and free T₄ levels during hGH therapy, and benefit from thyroid replacement. Most studies, however, indicate that children with normal thyroid function before treatment do not develop significant perturbations in thyroid hormone metabolism during hGH therapy.

Diabetes mellitus

The incidence of type 1 DM does not appear to be increased by hGH therapy, but some concerns regarding insulin resistance and increased risk for type 2 DM persist. Administration of hGH to non–GH-deficient prepubertal children is associated with increases in fasting insulin levels, mimicking (at an earlier than normal age) the increased hepatic glucose production and fasting insulin levels seen during normal puberty. Insulin levels typically rise, but return toward baseline when hGH is withdrawn, and euglycemia is maintained. Previous concerns about persistence of hGH-induced insulin resistance in patients with TS have been mitigated to some degree by awareness that decreased insulin sensitivity may be intrinsic to the disorder. Estimates of the incidence of type 2 DM among hGH recipients vary between US (14 per 100,000—not significantly different from general population rates) and international (34.4 per 100,000—roughly sixfold higher than reported in children not treated with hGH) databases. This difference may reflect differences between study populations with variable risk factors for the development of DM, and it is of some concern, as change in recipient characteristics due to ethnic demographics and rising childhood obesity rates could lead to an hGH-induced diabetes risk that differs from current estimates. The majority of patients who developed type 2 DM had preexisting risk factors for diabetes such as TS, PWS, and obesity, as well as having been born SGA and GHD due to leukemia and irradiation. Even in these patients at greater risk for disordered carbohydrate metabolism, hGH-associated diabetes is rare and it resolves in the vast majority of patients with discontinuation of the hGH.

Disproportionate growth

Concerns of disproportional growth have also risen in the context of children treated with higher-than-approved replacement doses of GH especially toward the final stages of treatment. IGF-1 levels in some patients can approach those found in acromegaly, and development of acromegalic features (eg, large hands and feet and coarse facial features) has been reported.

Malignancy concerns

Perhaps the greatest concern regarding hGH therapy is the theoretical possibility that it could facilitate the development of cancers. GH is mitogenic and IGF-1 inhibits apoptosis. Evidence from animal studies shows a cause-and-effect relation between supraphysiologic doses of hGH and development of leukemia. In humans, some, but not all, epidemiologic studies associate higher levels of IGF-1 with breast, colon, and prostate cancer. There are three clinical settings that have raised concern: (1) hGH-induced new malignancy; (2) hGH-induced recurrent malignancy; and (3) hGH-induced second malignancy in those already treated for one tumor. In 1988, reports from Japan describing leukemia in hGH-treated children raised concern about new malignancy. Subsequent analyses, based on much larger total patient-years of hGH treatment, indicate that the rate of new leukemia in (non-Japanese) patients without preexisting risk factors who are treated with hGH is not greater than expected for the general population. This lack of increased risk in children without preexisting risk factors has subsequently been confirmed in the Japanese population. Furthermore, the early concerns about de novo leukemia in patients without risk factors have not been substantiated by long-term postmarketing surveillance studies such as the National Cooperative Growth Study (NCGS) that monitored the safety and efficacy of hGH in 54,996 children older than 20 years.⁶⁰ In addition, new nonleukemic malignancies (intra- and extracranial) are not significantly increased by hGH treatment in patients without risk factors.

No report has associated hGH therapy with increased frequency of tumor recurrence. However, assessing the risk for recurrence of CNS tumors is particularly complicated due to lack of reliable knowledge about the natural history of this heterogeneous group of tumors. A cautious interpretation of current reassuring data is still appropriate. Most recurrences occur within the first 2 years of treatment, and most endocrinologists defer institution of treatment until a year of stable remission of a known tumor has passed. Though this may result in some lost growth for the child, it also avoids an assumption of causation between hGH therapy and tumor recurrence during this high-risk period. About second neoplasms, data thus far are consistent with a detectable increased risk associated with hGH therapy. Overall, the dominant risk factor for development of a second neoplasm appears to be a prior exposure to radiation. CNS tumors (most notably meningioma with a relative risk of 2.15⁶¹), followed by osteogenic sarcoma, have been the most frequently reported second neoplasms. Increased risk of developing second neoplasms in GH-treated childhood cancer survivors is now listed in US labeling for all hGH products. However, more recent and longer-term follow-up of subjects with CNS tumors treated with cranial irradiation show no difference in tumor recurrences or secondary tumors between those treated and not treated with hGH, raising questions about ascertainment bias in prior studies.⁶² Until this issue is resolved, patients who have developed GHD due to of treatment of malignancy and their families should be counseled about this possible adverse effect of hGH prior to commencing therapy.

Pseudotumor cerebri

Edema and sodium retention rarely occur early in the course of hGH therapy (particularly in older, heavier children, and adolescents), attributable to a direct antinatriuretic effect of hGH and/or IGF-1 on the renal tubule and minor elevations in plasma renin activity and aldosterone. Occasionally, fluid shifts within the CNS are sufficient to cause benign intracranial hypertension (pseudotumor cerebri) and its symptoms of headache, visual loss, vomiting, and papilledema. It is speculated that direct fluid-retaining properties of GH and/or action of locally produced IGF-1 on cerebrospinal fluid (CSF) production are causative. Most instances have occurred during early (though not invariably) treatment of patients with severe GHD or in the presence of other risk factors for this condition (eg, CRI and PWS). Cessation of GH therapy reverses symptoms, and resumption of hGH treatment can be successfully accomplished with reinitiation at a lower dose and gradual return to the initial dose. It is recommended that funduscopic examination be performed on all patients before initiation of GH therapy and periodically thereafter.

Slipped capital femoral epiphysis

hGH treatment causes an increase in both the proliferative and hypertrophied zones of the growth plate that may reduce the force needed to shear it. An increased frequency of slipped capital femoral epiphysis (SCFE) has been reported during treatment with hGH, but it is also associated with both hypothyroidism and (untreated) GHD. This rare complication is more common in children with organic causes of severe GHD, who tend to exhibit the most marked increase in growth rates with hGH replacement. These children require close monitoring for limp and hip or knee pain. Whether it is hGH therapy or simply rapid growth per se that plays a role in the development of SCFE is difficult to determine, particularly because the incidence of SCFE varies widely (2-140 cases per 100,000) depending on age, sex, race, and geographical location.

Adrenal insufficiency

In children with multiple pituitary hormone deficiencies, fatalities and other severe adverse events attributable to adrenal insufficiency (AI) have been reported during hGH therapy. In the NCGS report, the rate of AI (200 per 1 million) appeared to coincide with expected rates of secondary AI in the general population.⁶⁰ Nevertheless, GH affects the metabolism of glucocorticoids in ways that could theoretically increase risk for cortisol deficiency. Specifically, GH reduces activity of hepatic 11β-hydroxysteroid dehydrogenase 1, decreasing the conversion of cortisone to cortisol. GH also enhances the activity of CYP3A4, a cytochrome P450 enzyme that contributes to glucocorticoid catabolism. Thus, hGH therapy in a GHD patient may both unmask unsuspected ACTH deficiency and reduce the effect of low replacement doses of glucocorticoid. At the start of hGH treatment in these patients, adjustment of glucocorticoid replacement and stress dosing may be needed to sustain adequate cortisol coverage, and anticipatory guidance regarding signs and symptoms of AI is recommended.

Miscellaneous side effects

Concerns raised regarding other potential adverse effects of hGH therapy (eg, reduced testicular volume, gynecomastia, deterioration of renal function in patients with CRF, and cardiac ventricular hypertrophy in patients with Turner and Noonan syndromes) have, to date, either not been realized or considered to be of insufficient clinical significance to alter prescribing.

In summary, nearly two decades of experience with hGH has proven this therapy to be remarkably safe when used in conventional substitution doses for hGH deficiency. Higher-dose therapy for other indications also appears to be safe, but continued surveillance of its metabolic effects is indicated. There appear to be very few medical contraindications to hGH therapy. Nevertheless, cohort studies, from both the pituitary GH⁶³ and hGH⁶⁴ eras (based on a small number of cases), suggest that it is conceivable that mortality is affected by prior treatment. These data and the experience of transmission of CJD via pituitary GH are reminders that a farsighted view must be taken to study potential ramifications of hGH therapy.

Limited availability of pituitary GH once provided a barrier to expanding its use beyond children who were unequivocally GH deficient. Today, increased supply of hGH has been matched by increased demand. Thirty years ago in the United States, there was one approved indication for hGH, two hGH manufacturers, and approximately 10,000 children receiving treatment at a cost ranging between $5000 and $40,000 per patient per year. Today, we have 10 FDA-approved indications forhGH (see Table 2-5), multiple US and international manufacturers, and well over 50,000 active patients being treated just in the United States. Nevertheless, hGH therapy remains expensive. Relaxed diagnostic criteria have obliterated any clear boundary between GHD and GH sufficiency. This, along with FDA approval of hGH treatment for ISS, now allows a very large number of partially GH-deficient and non–GH-deficient children access to treatment.

Widespread distribution of hGH has been partially deterred by high drug costs. Prescribing hGH requires a difficult and often uncomfortable balancing of responsible use of medical resources with an obligation to do the best for each individual patient. Concern about psychological harm is invoked as a primary rationale for treating short stature yet data confirming both the association of short stature with psychological and emotional problems and efficacy of hGH therapy in alleviating the psychosocial consequences of short stature are scarce.⁶⁵ If the ultimate goal of hGH therapy is not tall stature but, rather, an improved quality of life, documentation of psychosocial impairment caused by short stature prior to therapy and improvement following hGH therapy ought to play an important role in the initiation of GH therapy and evaluation of its efficacy. To date, however, growth rate and adult height remain the measures by which therapeutic success is judged. Lack of definable quality-of-life benefit makes it challenging to assess the risk-to-benefit ratio in counseling and obtaining informed consent from families and patients. Though hGH treatment has a strong safety record, it is costly, requires daily injections, and even a very low risk for remote adverse effects is relevant to treatment decisions in an otherwise healthy child.³

Determining an appropriate endpoint for hGH therapy for short stature remains controversial. Attainment of genetic potential for height, a realistic possibility with optimal diagnosis and treatment, remains a goal for many. Consistent adherence to the goal of alleviating disabling short stature, on the other hand, implies that hGH therapy for stature (ie, not for GH replacement in severely GH-deficient individuals) be discontinued when each treated child reaches an adult height within the normal adult distribution.⁶⁶

Tall stature and excessive growth disorders are relatively rare occurrences in pediatric endocrinology; however, these disorders may also warrant careful evaluation. Primary growth disorders may have prenatal onset and are caused by an intrinsic defect in bone and/or connective tissue and often are of genetic origin. Growth in these states is frequently disproportionate and dysmorphic features are common. Secondary excessive growth syndromes are, in general, the result of hormonal disturbances. Pathologic conditions can present with excessive growth, reflecting effects of intrauterine factors, genetic factors, and hormonal stimulation of growth, including GH excess.

Physiologic Variations

Intrauterine overgrowth

An infant born with a length and weight above the 90th percentile for gestational age (excessive intrauterine growth) is defined as “large for gestational age” (LGA). This condition is commonly caused by maternal DM and should prompt maternal evaluation even in the absence of a positive clinical history. In the postnatal period, blood sugars should be carefully monitored as intravenous infusions of glucose may be necessary to prevent severe insulin-induced hypoglycemia. In the present obesity era, however, prepregnancy overweight/obesity is the most common risk factor for an LGA birth.⁶⁷

Constitutional tall stature

As with constitutional short stature and delay in maturation, constitutional tall stature is a variant of normal childhood growth. By definition, height is between 2 and 4 SD above the mean height for age. Birth length is usually normal, but accelerated growth during the first 12 to 15 months of life leads to upward crossing of percentiles toward a genetically determined (usually the mid-parental height) height percentile. Continuation of this trend between 15 and 30 months can occur if a familial tendency toward early puberty and/or overnutrition is present. Tall stature is evident by age 3 years, and subsequent growth is usually parallel to, but above the 95th percentile. Body proportions are normal, and the timing of puberty reflects familial characteristics. Genetic and familial factors, often from both parents, can explain the growth pattern of most children with tall stature. GH secretion and levels of IGF-1 and IGFBP-3 can be (but are not invariably) in the high-normal range, and elevated ratios of IGF-1/IGFBP-3 suggest greater bioavailability of IGF-1 for target tissues.

Because of societal definitions of normal height, cultural acceptance of tall stature and lack of parental concerns over a tall child cloud the true “incidence” of this growth pattern. Fortunately, tall stature in females has become less stigmatizing and women are finding value in athletic and professional opportunities in which tall stature confers an advantage. Traditionally, treatment to attenuate height has relied on administration of gonadal steroids to hasten epiphyseal maturation. Discussionabout therapy should be reserved for extreme cases (eg, boys with predicted height > 203 cm = 80 in and girls > 186 cm = 73 in). High-dose estrogen treatment (100-300 μg of ethinyl estradiol) and high-dose testosterone treatment (250-1000 mg/month) have been used to reduce final height of tall girls and boys, respectively. An alternative option is to induce early puberty with low doses of sex steroids. Potential risks must be weighed against uncertain clinical benefit. A dose-response relationship between fertility in later life and estrogen dose used for the treatment of tall stature in adolescent girls has been documented. A higher estrogen dose is associated with an increased rate of infertility.⁶⁸

Primary Excessive Growth Disorders

These include chromosomal disorders such as Klinefelter syndrome (47,XXY), 47,XYY, genetic syndromes such as Beckwith-Wiedemann syndrome, Sotos syndrome, Marfan syndrome, Simpson-Golabi syndrome, and others.

Klinefelter syndrome

Klinefelter syndrome (KS) results from the presence of two or more X chromosomes in a male; 47,XXY is most common, with mosaic karyotypes (eg, 46,XY/47,XXY) and other aneuploid karyotypes (eg, 48,XXXY) occurring much less frequently. It occurs in between 1:500 and 1:1000 live male births and is highly underdiagnosed, with approximately 10% of cases diagnosed before puberty. The presence of an extra X chromosome leads to testicular hyalinization and fibrosis, impaired testosterone production, and infertility. Infants with KS are rarely diagnosed because they lack disease-specific stigmata. Rarely, depending on the severity of fetal testicular hypofunction, genital abnormalities such as unusually small testes, microphallus, cryptorchidism, and hypospadias may be present in the newborn period. Other clinical manifestations include abnormal body proportions (ie, long legs and low U/L segment ratio), mild mental retardation (reduced verbal IQ) and/or behavioral difficulties, learning disabilities, delayed or (more commonly) incomplete puberty, and gynecomastia. The clinical diagnosis is confirmed by demonstration of an extra X chromosome in the karyotype. It is more doubtful whether parental origin of the extra X chromosome influences phenotype. One of the extra X chromosomes should undergo inactivation as in females; however, the phenotype is presumed to be secondary to noninactivated extra genes on the X chromosome, although other genetic mechanisms are possible. The SHOX gene situated in the pseudoautosomal region 1 on the short arm of the X chromosome has been shown to influence the accelerated growth in patients with KS. The X chromosome carries genes that are expressed in testis function, brain development, and growth, and, therefore, is likely to play a role in KS.

Treatment of patients with KS requires a multidisciplinary team, including pediatricians, endocrinologists, urologists, infertility specialists, and psychologists. Treatment options in adolescents and adults differ. The loss of spermatogonial cells occurs progressively, and most boys with KS undergo massive apoptosis during puberty. Laboratory evidence for hypergonadotropic hypogonadism becomes manifest, with elevations in FSH levels typically exceeding those in LH, and testosterone levels that are low for age during puberty and rarely spontaneously rising above the low-normal adult male range. It should be noted, however, that most patients with KS demonstrate transition to gonadal puberty without reliance on exogenous androgens.⁶⁹ When androgen deficiency and evidence for testicular failure are severe, gradual introduction of a long-acting intramuscular testosterone ester to a full replacement dosage of 200 to 300 mg every 2 weeks will restore testosterone levels to the normal range. As it appears possible that exogenous androgen treatment may negatively affect chances for later sperm extraction, this should be discussed with the family. Other treatment options include transdermal gel preparations that offer the best pharmacodynamic profile, but intramuscular injections remain popular due to ease of administration. The aim of treatment should include normalization of serum testosterone levels into the mid-normal range.⁷⁰ Gynecomastia (caused by aromatization of exogenous or endogenous androgens) can be severe, and it may require surgical therapy, although treatment with an aromatase inhibitor at this stage has been suggested. Fertility preservation should be discussed in patients with KS duringpuberty noting that there is opportunity for cryopreservation of sperm. Recent studies have demonstrated residual foci of spermatogenesis in testicular biopsies of prepubertal boys with KS despite the characteristic extensive fibrosis and hyalinization of the seminiferous tubules. Sperm found in the testes of men with KS have only a slight increased frequency of sex chromosome abnormalities. The introduction of novel techniques (testicular sperm extraction [TESE] in combination with intracytoplasmic sperm injection [ICSI]) has allowed KS males to father children.⁷¹

47,XYY

The presence of an extra Y chromosome, which occurs in between 1:500 and 1:1000 live male births, also predisposes to tall stature. The incidence is markedly increased in males whose height exceeds 183 cm (72 in). Other reported findings in affected individuals include hypospadias, cryptorchidism, severe acne, and variable degrees of mental retardation. However, broader studies indicate that prediction of behavioral problems cannot be accurately made based on the finding of a 47,XYY karyotype.

Beckwith-Wiedemann syndrome

Beckwith-Wiedemann syndrome (BWS) is a rare cause of intrauterine overgrowth with a reported incidence of approximately 1 in 13,700 live births. Hypoglycemia frequently occurs and is secondary to islet cell hyperplasia causing hyperinsulinemia. Clinical features include excessive somatic and organ overgrowth, hemihypertrophy, macroglossia, umbilical abnormalities (omphalocele, umbilical hernia, and diastasis recti), craniofacial abnormalities (mid-face hypoplasia, prominent occiput, flat nasal bridge, and high-arched palate), earlobe abnormalities, renal abnormalities, and embryonal tumors (eg, Wilms tumor, hepatoblastoma, neuroblastoma, rhabdomyosarcoma, and adrenocortical carcinoma). Overexpression of the paternally derived gene for IGF-2 may account for fetal overgrowth, but no postnatal abnormality of the GH-IGF axis has been identified. Approximately 85% of reported BWS cases are sporadic, whereas the remaining 15% are familial. BWS is a multigenic disorder, resulting from dysregulation of several closely linked genes associated with cell cycle and growth control on chromosome 11p15. Imprinted genes implicated in the etiology of BWS include the paternally expressed genes, IGF2 and KCNQ1OT1, and the maternally expressed genes, H19, CDKN1C, and KCNQ1. Normally, the H19 and IGF2 genes are coordinately regulated so that H19 is only expressed from the maternal allele whereas IGF2 is only expressed from the paternal allele. Several molecular abnormalities are associated with BWS. The largest molecular subgroups of BWS (60%-70% of cases) have imprinting defect on chromosome 11p15. Approximately 15% of the patients may also have paternal uniparental disomy (two paternally derived copies of 11p15 and no maternal contribution for that region). Five to 10 percent of individuals with BWS carry point mutations in CDKN1C and 2% to 3% of cases have chromosomal rearrangements. Finally, mutations outside the chromosome 11p15.5 critical region have also been identified in patients with BWS. Various alterations in the growth regulatory genes in the 11p15 region may also account for somatic asymmetry and the association of BWS with increased risk for various cancers.

Sotos syndrome

Children with Sotos syndrome are also usually born LGA and demonstrate abnormally rapid growth during the first 3 to 4 years of life. Typical facial features include sparse frontotemporal hair, high bossed forehead, downslanting palpebral fissures, a long narrow face, and prominent narrow jaw; the head is said to resemble an inverted pear. The facial shape is retained into adulthood, but with time the chin becomes squarer in shape and more prominent. Other phenotypic features include enlarged hands and feet with thickened subcutaneous tissue. Body proportions are abnormal with arm span exceeding height by as much as 5 cm. Approximately 90% of children have a height and/or head circumference two or more SD above the mean. Growth velocity remains rapid in childhood, but with advancement in BA and early puberty resulting in early maturation of the growth plates resulting in a final height in the normal range. Macrocephaly is usually present at all ages. Affected children also have a nonprogressive neurologic disorder associated with mental retardation and delayed motor milestones. It is an autosomal dominant disorder although most cases with Sotos syndrome arise de novo. Sotos syndrome is caused by a deletion or mutation in the NSD1 gene, which maps to 5q35. The NSD1 gene encodes nuclear receptor set domain protein 1, which enhances androgen receptor transactivation. In approximately 10% of classic cases of Sotos syndrome, NSD1 abnormalities are not identified.

Marfan syndrome and related disorders

Marfan syndrome is an autosomal dominant disorder of collage metabolism characterized by hyperextensible joints, upward and outward dislocation of the lens (ectopia lentis), kyphoscoliosis, and aortic root dilatation. Height is increased, and body proportions are abnormal owing to excessive arm and leg length. Arm span is greater than height, and the U/L segment ratio is diminished. Long fingers and toes (arachnodactyly), pectus excavatum, and scoliosis are observed in most patients. Although historically it has been considered as a condition caused by mutations in the fibrillin gene located on chromosome 15, the current understanding has shifted to a model that involves dysregulation of cytokine-transforming growth factor-β (TGF-β) signaling.⁷² Loeys-Dietz syndrome is caused by mutations in either of two TGF-β receptor (TGFBR1 or TGFBR2) genes. Affected patients share many of the features of Marfan syndrome except for lens dislocation. Life expectancy is reduced primarily because of cardiovascular manifestations such as aortic regurgitation and dissection.

A similar body habitus and tendency toward tall stature is observed in patients with homocystinuria, an autosomal recessive disorder caused by deficiency of cystathionine β-synthase. This enzyme catalyzes the combination of homocystine and serine to form cystathionine; a deficiency leads to elevations of homocystine in blood and urine. Patients usually appear normal at birth, but mental retardation, downward and inward dislocation of the lens, and marked myopia and/or astigmatism are noted in early childhood. Osteoporosis, scoliosis, and thromboembolic phenomenon (arterial and venous) can also occur. Treatment includes dietary methionine restriction and, in responsive patients, pyridoxine administration. In some states, newborn screening for this rare disorder facilitates very early diagnosis.

Secondary Growth Excess Disorders

Growth hormone excess

GH excess is a well-described but rare cause of tall stature and rapid growth in childhood. The etiology is usually a functioning pituitary adenoma driven by a constitutively activated G-protein reducing GTPase activity, increasing cAMP, and promoting excess GH secretion. Though usually an isolated finding, GH-producing tumors can occur in association with McCune-Albright syndrome (multiorgan G-protein activation), tuberous sclerosis, Carney complex, and familial isolated pituitary adenoma (FIPA). GH excess may also result from unregulated GH secretion as in neurofibromatosis. When prolonged GH excess occurs prior to epiphyseal fusion or in the context of delayed puberty or absent gonadotrophs, adult heights typically exceed MPH and can be well above the normal range.

Diagnosis is suggested by accelerated growth diverging from the expected familial pattern. Skeletal age is normal or mildly advanced, yielding excessively tall adult height predictions. Basal serum GH levels may be normal or increased; given the pulsatile nature of normal GH secretion, reliance on random samples can lead to both false-positive and false-negative impressions. Additional information suggesting the diagnosis includes lack of suppression of GH levels in response to an oral glucose load. Normally, an oral glucose load prompts insulin secretion, suppression of IGFBP-1 resulting in an increase in free IGF-1 levels, and subsequent negative feedback on GH release. IGF-1 and IGFBP-3 levels are useful screening tests given their correlation with 24-hour GH secretion but are not always present in excess of values expected in normal puberty. Suggestive clinical and laboratory findings should prompt imaging of the hypothalamic/pituitary region with a contrast-enhanced MRI.

Treatment is aimed at restoring the GH/IGF-1 axis to normal while preserving function of other pituitary hormones. Transsphenoidal resection can be curative for well-circumscribed GH-producing adenomas, but a transcranial approach may be necessary for more extensive lesions. Adjunctive medical and radiation may be indicated for tumors not completely resectable. The effects of radiation are typically delayed by several months to years (ie, GH levels decreased to 50% of initial levels by 2 years, 75% by 5 years, and 90% by 15 years), and loss of other pituitary function is common. Somatostatin analogs (octreotide) can be used to suppress GH levels and normalize IGF-1 as an alternative to surgical or radiation treatment. In children, octreotide is given by daily subcutaneous injections. Continuous subcutaneous pumps have also been used. Long-acting preparations are available and have been effective in adult patients. Bromocriptine, a dopamine agonist, also leads to the reduction in GH secretion, but most patients do not achieve normal IGF-1 levels and side effects (nausea, abdominal pain, fatigue, and orthostatic hypotension) are common. In addition, an analog of GH that functions as a GHR antagonist (pegvisomant) has been shown to be effective and is approved for use in adults to treat GH excess caused by pituitary tumors or ectopic GHRH hypersecretion. Long-term and pediatric studies are needed.

Estrogen resistance and aromatase deficiency

Very tall adult stature secondary to hormonal disorders has also been described in rare cases with estrogen resistance due to mutations in the ER gene and secondary to mutations in the aromatase gene (CYP19). In males, the former is associated with elevated serum levels of estradiol and estrone, whereas in the latter, serum levels of estradiol and estrone are very low; in both situations, there is little or no estrogen effect, due to resistance and deficiency, respectively. Though sexual development, puberty, and sexual orientation are normal, BA becomes markedly delayed, epiphyses remain open, and there is absence of the pubertal growth spurt with continuation of growth into adulthood. In addition, there is reduced BMD and osteoporosis in both sexes. In females with aromatase deficiency, secondary sexual maturation fails to occur whereas during puberty, virilization and multicystic ovaries develop because of elevated gonadotropins and androgens. Placenta-specific aromatase deficiency results in virilization of the mother and her female fetus because of the accumulation of potent androgens that are not converted to estrogens.

Unwanted growth and growth restriction

Premature exposure to sex steroids, particularly estrogen, can attenuate linear growth potential by causing accelerated skeletal maturation, thereby limiting the linear growth achieved before growth plate closure. Children with untreated central or peripheral precocious puberty show growth rate acceleration, with rapid BA advancement, early cessation of growth, and eventual short stature. Low concentrations of estrogen stimulate predominantly ER-α to enhance chondrocyte growth, whereas high concentrations stimulate ER-β to inhibit cartilage cell division and promote maturation and senescence of chondrocytes. Because of its ability to advance skeletal maturation in excess of normal height age increases, estrogen has been historically used as a growth-attenuating agent in females for whom the prospect for excess height was viewed as a disability (see the preceding discussion). During the past 30 years, however, increased social acceptance of tall stature in females and the greater visibility of positive female role models who are tall have virtually eliminated requests for such treatment. Furthermore, estrogen treatment of tall females may not be without risk, including diminished fertility with longer exposure periods.⁶⁸ Information regarding growth attenuation treatment for tall-statured boys is not available.

More recently, growth attenuation with high-dose estrogen for children with profound physical and cognitive disabilities has been reported. The rationale for such treatment includes the potentially negative effect of increasing size of a severely disabled and nonmobile child on the ability of his or her family to provide independent care in the home and access to life outside the home. Assessing the effectiveness of such treatment for profoundly disabled children is complicated by preexisting growth-compromising problems (eg, compromised bone mineralization and scoliosis). Importantly, “ease-of-care benefit” resulting from growth attenuation assumes that weight reduction will be commensurate with length reduction. Current approaches include provision to severely disabled girls of exogenous estrogen orally or by transdermal patch. In addition, possible benefits of growth attenuation could mean that nontreatment of spontaneous precocious puberty (more common in these children) should also be discussed as a therapeutic alternative with long-term ease-of-care benefits that might outweigh the shorter-term advantages of suppressing pubertal progression. Experience to inform the safety and efficacy of growth attenuation therapy for children with profound cognitive disability is limited, so that it should still be viewed as innovative therapy that offers the possibility of an improved quality of life. Discussion of this option would optimally occur as part of anticipatory guidance at approximately the age of 3 years so that the potential benefits can be maximized. Ethics consultation is recommended.⁷³