Growth Hormone Deficiency: Diagnostic Principles and Practice

Transcription

1 Ranke MB, Mullis P-E (eds): Diagnostics of Endocrine Function in Children and Adolescents, ed 4. Basel, Karger, 2011, pp Growth Hormone Deficiency: Diagnostic Principles and Practice Michael B. Ranke University Children s Hospital, Tübingen, Germany Growth hormone deficiency (GHD) by and large involves a wide range of disorders, including an impairment in GH secretion and/or a reduced impact of GH (table 1) [1]. It has been proposed that GH and IGF- 1 are analogous to other endocrine systems in that they are interconnected, with GH being the regulatory hormone and IGF- 1 the hormone acting primarily. A distinction is thus made between primary IGF deficiency, which involves IGF- 1 impairment but normal GH secretion and action, whereas secondary IGF deficiency is a state in which IGF- 1 is impaired due to irregularities in GH secretion [2]. These disorders are diagnosed via a combination of diagnostic tools [see also chapters by Blum and Albertsson- Wikland, this vol.]. In addition, imaging techniques and molecular genetic methods [see chapter by Pandey and Mullis, this vol.] can be applied to confirm and/or specify the diagnosis. More specifically, GHD is characterized by GH secretion levels that are below those for age- and sex- matched norms. In a broader sense, GHD is the clinical disorder resulting from the impaired secretion or action of GH in an individual. The first task in the diagnostic procedure should thus be to establish whether or not the clinical presentation of an individual corresponds with the situation of impaired GH secretion or action. If this is plausible, evidence must be sought, in a second step, for an impairment in GH secretion (or action). The difficulty in associating an individual s anthropometrical and biochemical phenotype with GH impairment is partly related to the fact that the phenotype may vary, depending on age and severity, as well as according to the duration of GH impairment at the time the patient presents. Moreover, certain signs or symptoms typical for GHD are also attributable to other causes. A combination of typical signs could thus be more helpful than a single characteristic specific to GH impairment. Finally, it must be mentioned that the empirical basis used to make a causal association between the quantitative deviation of anthropometrical/biochemical signs and a measurable degree of GH impairment is very narrow.

2 Table 1. Disorders of the GH- IGF axis (Wit et al. [1]) ESPE code Diagnosis 1B.3 growth hormone deficiency (secondary IGF deficiency) 1B.3a 1B.3a.1 1B.3a.2 congenital GHD associated with other complex syndromes known genetic defects 1B.3a.2a 1B.3a.2b 1B.3a.2c 1B.3a.2d 1B.3a.2e 1B.3a.2f 1B.3a.2g 1B.3a.2y -HESX1 -LHX3 -LHX4 -PROP1 -POU1F1 -GHRHR -GH - other specified genetic defects 1B.3a.3 1B.3a.4 1B.3a.9 associated with cerebral or facial malformations (e.g. SOD, empty sella syndrome, solitary central maxillary incisor syndrome, mid- line palatal cleft syndrome, arachnoidal cyst, congenital hydrocephalus, HME = hypoplastic anterior pituitary+missing stalk+ectopic posterior pituitary) associated with prenatal infections Idiopathic GHD 1B.3a.9a 1B.3a.9b classical idiopathic GHD GH neurosecretory dysfunction 1B.3b acquired GHD 1B.3b.1 1B.3b.2 1B.3b.3 1B.3b.4 1B.3b.5 1B.3b.6 1B.3b.7 1B.3b.8 -craniopharyngioma - other pituitary tumors - cranial tumors distant to pituitary - tumors outside the cranium -head trauma -CNS infection -granulomatous disease -vascular anomaly GHD Testing 103

3 Table 1. Continued ESPE code Diagnosis 1B.4 other disorders of GH- IGF axis (primary IGF- 1 deficiency and resistance) 1B.4a 1B.4b 1B.4c 1B.4d 1B.4e 1B.4f 1B.4z bio-inactive GH anomalies of GH receptor anomalies of GH signal transduction ALS deficiency IGF-1 deficiency IGF- 1 resistance (IGF- 1R defect, postreceptor defect) other disorders other causes of disorders of the GH- IGF axis 1B.5 endocrine disorders 1B.6 metabolic disorders 1B.7 psychosocial disorders 1B.8 iatrogenic The problem faced by the clinician in diagnosing GHD has been examined from different levels [3 8]. 1 The first difficulty is in defining the diagnostic entrance criteria in such a way that the secondary diagnostic steps can thereby be justified and not excluded. If very stringent criteria (e.g. cut- offs) are a constituent part of the initial diagnostic steps, some individuals may possibly be excluded from the further work- up. This is an error of the first order (although the patient is affected, she/he is defined as healthy). On the other hand, the criteria must be chosen in such a way as to avoid an error of the second order (the individual is considered sick, and therefore subjected to further possibly unnecessary diagnostic procedures; however, she/he is in fact healthy). 2 In a second step, proof of an impairment in GH secretion must be obtained. The strain on the patient, arising from the scope of the diagnostic procedure, should be limited and errors of the first or second order should be minimized. 3 In a third step, techniques to define the causes of impaired GH secretion need to be applied. 4 Finally, there is the question about whether the altered signs and symptoms resulting from the normalization of GH through replacement therapy can be used to confirm or refute the diagnosis of GH impairment, and thus become part of a secondary diagnostic step [9, 10]. 104 Ranke

4 Table 2. Criteria for estimating the likelihood of GHD before GH testing Congenital GHD infancy Congenital GHD childhood Acquired GHD likely less likely likely less likely likely less likely Perinatal history hypoglycemia ikterus prolonged traumatic birth family history of GHD hypoglycemia ikterus prolonged traumatic birth family history of GHD head trauma postmalignancy neurosurgery meningitis Phenotype doll-like n.r. n.r. Age, years n.r. m > 12 f > 10 n.r. Ht n.r. <-2.5 SDS >-2.0 SDS n.r. Ht - MPH n.r. >- 1.3 SDS <- 1.3 SDS n.r. Wt Wt SDS < < Ht SDS Wt SDS - Ht SDS > Wt SDS < < Ht SDS Wt SDS < < Ht SDS Ht velocity <- 1.0 SDS >- 1.0 SDS <- 1.0 SDS >0.0 SDS <- 1.0 SDS >0.0 SDS Bone age, years n.r. CA- BA >> 1.0 BA > = CA n.r. Head circumference (HC) <- 2.0 SDS n.r Ht SDS HC SDS n.r. Body proportions n.r. SH SDS > > Ht SDS n.r Other pituitary deficit yes n.r. yes n.r. yes no IGF-1 <-1.0 SDS >-1.0 SDS <-2.0 SDS >-1.0 SDS <-2.0 SDS >-1.0 SDS IGFBP- 3 <- 2.0 SDS >- 1.0 SDS <- 2.0 SDS >0.0 SDS <- 2.0 SDS >0.0 SDS BA = Bone age; CA = chronological age; GHD = growth hormone deficiency; HC = head circumference; Ht = height; MPH = mid- parental height; n.r. = not relevant; SDS = standard deviation score for CA; Wt = weight. Considerations before GH Testing Entrance Criteria It is important to ensure that a combination of various anamnestic, auxological and biochemical information is available before GH testing is done in infants, children and adolescents with GHD, as this is crucial to establishing the diagnosis of GHD firmly (table 2). The proper classification of a disorder on the GH- IGF axis must be based on biochemical analyses (table 3). GHD Testing 105

5 Table 3. Pattern of biochemical findings in varying disorders of growth hormone deficiency (GHD) syndrome Diagnosis (level of disorder) GH (standard tests) GH (GHRH test) GH (spontaneous secretion) IGF-1 IGFBP-3 Comment Primary disorders of GH- IGF axis Disorders related to GH deficiency GH insufficiency (GH gene deletion) GH insufficiency (pituitary level) GH insufficiency (hypothalamic level) GH insufficiency (bioinactive GH) GH insufficiency (neurosecretory dysfunction) potentially growthinhibiting GH- ABs to GH treatment /n growth response to GH therapy n /n therapy with GHRH possible n n n growth response to GH therapy n /n /n growth response to GH therapy Disorders related to impaired action of GH or IGF GHBP excess/ antigh ABs States of congenital GH impaired IGF- 1 generation (primary IGF deficiency) Primary IGF resistance (IGF receptor/postreceptor defect) n/ n/ GHBP measurement, IGF generation test growth improves with exogenous GH therapy with IGF- 1 possible no therapy known to date Secondary disorders of GH- IGF axis Acquired inhibition of IGF- 1 action (e.g. IGFBP excess in chronic renal failure) (CRF) /n /n /n /n growth improves with exogenous GH Hyperalimentation n n/? n tall stature 106 Ranke

6 Table 3. Continued Diagnosis (level of disorder) GH (standard tests) GH (GHRH test) GH (spontaneous secretion) IGF-1 IGFBP-3 Comment Hypercortisolism, hypothyroidism Diabetes mellitus, fasting, hepatopathy /n n no therapy with GH n = Normal; = increased; = diminished. History Genetic forms of isolated or combined GHD are rare and only occur in about 3% of cases [11, 12]. Thus, in addition to measuring height in both parents, a detailed family history needs to be recorded. Other historical events, such as head trauma [13] or treatment for malignancies [14, 15] are also strong indicators of GHD in a short child. Perinatal History and Signs In the early years of research on GHD, it was observed that a high proportion of affected children were delivered vaginally and in a breech position [16]. Whether GHD is caused by a trauma possibly occurring during such a delivery has yet to be clarified. Evidence has also to be found for pituitary defects in the fetus which possibly led to a breech position. It is standard practice to perform a caesarean section in most cases in which an unfavorable presentation at birth is apparent. Nevertheless, a head trauma during delivery may cause pituitary damage similar to postnatal head trauma [13]. A small penis and bilateral undescended testes can be an indication of insufficient fetal exposure to gonadotrophins/testosterone in males. This may be associated with GHD. Prolonged jaundice may also be a symptom of GHD combined with hypocortisolism [17]. Hypoglycemia may occur in congenital GHD [18]; however, it is not clear whether this is possibly the result of isolated GHD or, instead, of GHD combined with other pituitary defects. In these instances GHD can be suspected even if small body size (length) is not a presenting symptom. Phenotypic Appearance Children with severe GHD have a relatively large neurocranium in relation to the mid- facial region. This gives them a typical doll- like appearance. Head circumference is about 2 SDS above height SDS [19]; thus, microcephaly is a very unlikely feature of GHD. Typical characteristics include small hands and feet and a less mature GHD Testing 107

7 voice, and the overall appearance does not correspond with age. Frequently, however, the appearance of children with GH is not conspicuous. Anthropometric Symptoms at Presentation The most frequent sign of GHD in a child is height development [20, 21]. Short stature is defined as height/length below the 3rd centile for age. Children with GHD are less likely to present with a height of > 2.5 SDS [22]. In addition, the child s height/ length needs to be examined in relation to parental height [23]. It is uncommon for a child s height to be within the range of the reference population but outside the target range for the parents (roughly: height SDS minus MPH SDS < 1.3 SDS), and further diagnostics are required to identify GHD. Since shortness in children with GHD is proportional (sitting height and arm span deviate to the same degree as does height), these anthropometrical parameters need to be documented and compared with appropriate references. Another important aspect to consider is height within the context of delayed puberty [24], as this is an explicit indicator of GHD. The height of children with pubertal age but prepubertal stage may be compared with extrapolated references based on the ICP model devised previously [25 27]. Since growth is not static, like height, but takes place in a dynamic process, height velocity is considered the more appropriate parameter for defining abnormal growth. Duche et al. [8] observed a sensitivity of 92%, with a specificity of 72%, in diagnosing GHD by combining a height velocity of < 1.0 SDS with an IGF- 1 SDS of < 2.0 SDS. However, the difficulty in applying height velocity is that two measurements must be taken at appropriate intervals in order to obtain an exact calculation. The time frame for the two measurements must include a stage of height velocity (the lower, the longer) and the calculation depends on the precision of the measuring device. Usually, data relating to 6 (infancy) to 12 (childhood) consecutive months are needed to establish accurate height velocity. If the diagnosis of GHD is likely in a given child, the diagnostic (therapeutic) process should not be delayed for the sake of height velocity, whereas a delay would be justified if the diagnosis of GHD is less likely. The question about which limits should be considered as abnormal in a short child is still controversial. A child with normal size at birth, who however grows slowly, will obviously have short stature. It is not clear what height velocity can be expected once height is below the normal range. If growth dynamics are expressed in terms of delta height (SDS), then a value below 0.5 SDS is congruent with normal variation. If growth dynamics are expressed in terms of height velocity (cm/year), then a value above the 25th centile for age (equal to approximately 1.0 SDS) is not very likely in a child with GHD, and a value above the 50th centile for age (approximately 0.0 SDS) is highly unlikely in GHD. In children at pubertal age but delayed puberty, special references need to be considered [28]. It is a common assumption that bone age (BA) is delayed and BA progression is retarded in GHD. However, there is considerable overlap in normal or short non- 108 Ranke

8 GHD children [29]. This may be due to the following reasons: (1) bone age determinations are prone to subjective error, (2) appropriate, contemporary references are usually not available, (3) there is a natural variation in BA at a given chronological age, and (4) bone age is not completely independent of height, with smaller children in an age cohort presenting with lower BA. Nevertheless, a bone age equivalent to or higher than chronological age is not likely in GHD. In GHD, weight is typically not diminished to the same degree as is height [30] (weight SDS minus height SDS > +1.0 SDS). The children look relatively wellnourished. However, overt obesity is the exception in congenital GHD. Although the deficit in muscle mass and the relative abundance of subcutaneous fat can be documented in GHD by means of special techniques [31], body composition assessment is at present not considered to be a standard component in the diagnostics of short stature. The same is true for bone mass determinations [32]. Since these aspects are significant in treating patients with permanent GHD during the transition to adult age, it is probable that pediatric practices relating to diagnosis and follow- up will be adapted for adult care in future. Biochemical Markers of GHD It is a well- known fact that various circulating biochemical compounds depend on GH secretion. Indicators of bone growth (e.g. alkaline phosphatase) and collagen metabolism are low in GHD patients, but normalize during GH therapy [see also article by Crofton, this vol.]. Particular mention must be made of the GH- dependent factors in the IGF system (e.g. IGF- 1, IGFBP- 3) whose role in the diagnostics of GHD in children and adults has been a controversial matter for the past few decades. The main issues pertain to the partial effects of both GH and IGF- 1 at the sites of growth as well as to the relevance of circulating IGF- 1 during growth [33 35]. A more specific discussion focused on the diagnostic role of IGF- 1 (and IGFBP- 3) blood levels in GHD. IGF measurements are advantageous because they can easily be obtained from a single sample; in addition, reference levels for children and adolescents are available [see chapter by Blum]. It must be remembered that this debate on the diagnostic role of IGFs is related to the fact that a golden standard for establishing the diagnosis of GHD does not exist. Authors have associated the diagnosis of GHD with the degree of sensitivity and specificity of IGF- 1 (IGFBP- 3) by means of their own definitions and diagnostic settings [36]. Interim conclusions show that low levels (in relation to age and sex) of IGF- 1 (IGFBP- 3) support the diagnosis of GHD, whereas normal levels (e.g. IGF- 1 > 1.0 SDS; IGFBP- 3 > 0.0 SDS) do not [37, 38] (table 2). Thus, IGF measurements have become an essential part of the work- up for GHD. Pituitary Imaging Special cerebral imaging techniques, such as computed tomography (CT) and magnetic resonance imaging (MRI) were conventionally done after the diagnosis of GHD GHD Testing 109

9 was established, with the aim of ascertaining the cause of GHD. A high frequency of abnormalities in the hypothalamus- pituitary region was documented by means of MRI in children with GHD [39]. A strong but not definite indicator of GHD could be an impairment in the size of the anterior pituitary, if MRI shows no evidence of a structural abnormality in the pituitary region [40]. The ready accessibility of imaging tools has led to a growing demand for techniques providing immediate results; however, although these methods are non- invasive, they are not necessarily suitable for examining younger children. In addition, operating the equipment and interpreting the results require considerable skill. A simple X- ray may often be more helpful in the further diagnostics of short stature in a child; e.g. calcifications in the pituitary region have explained 90% of craniopharyngioma cases. Growth Hormone Natural Variation of GH Molecules GH is the gene product of the GH- N gene which is primarily expressed in the somatotrophs of the pituitary but also, to a lesser degree, in other tissues [41, 42]. The major gene product (approximately 80 90%) is a peptide hormone of 191 amino acids (size 22 kda); however, due to alternative splicing, a shorter form (lacking amino acids 32 46; size 20 kda) is also produced, which accounts for 10 20%. Shorter and larger forms are also formed in the circulation, either by means of proteolytic fragmentation or as a result of an association of forms of GH. Thus, there is not only one natural GH, and the circulating amount of the different components may depend on their production, metabolic circumstances and varying degrees of degradation [43]. In the human placenta, a GH- variant gene (GH- V) is expressed. Since three different variants of the GH receptor exist, which, through conformational changes during GH binding, transmit signals to the target cell [44], it is conceivable that the different forms of circulating GH exert different effects [45]. GH forms also vary, depending on GHsecretory stimuli [46]. GH-Binding Protein There is an additional level of complexity arising from the fact that the external part of the GH receptor can be cleaved into the circulation by proteolysis and can bind GH [47]. It is assumed that up to 50% of circulating GH is complexed to this GH- binding protein (GHBP) [48, 49]. GHBP prolongs the half- life of GH by protecting it from degradation. The GH- GHBP complex is a hormone reservoir. Several reports refer to immunometrical assays used to determine GHBP [50 53]. The assays have not been standardized, and there is no international reference preparation available for GHBP. However, measurements in normal adults show that serum GHBP are relatively constant (ranging between approximately 0.5 to 3.8 nmol/l) and correlate positively with BMI and negatively with age. 110 Ranke

10 Central nervous system Hypothalamus Hypophysis Plasma Neurotransmitter GHRIH GHRH (?) GH IGF-producing tissues (e.g. liver) IGFs (?) Fig. 1. Model of the GH regulatory system. Growing tissue IGFs GHBP levels are low during fetal life, rise in the first years after birth and thereafter achieve adult levels [54 57]. The levels also correlate with BMI and are thought to reflect the individual sensitivity to GH [54]. In patients with primary IGF deficiency, GHBP is absent/diminished in serum [58] due to a defect in the extracellular domain of the GH receptor (GH resistance of the Laron syndrome type). On the other hand, GH action can be demonstrated by the excess of GHBP [59, 60]. It is probably relevant to measure GH if GHBP is absent [61] in these and other specific situations. Both GHBP and human antibodies (AB) to GH in the circulation may affect the reading of GH in an immunoassay. Human anti- GH AB bind part of the labeled GH within the assay, and the amount of label bound to the non- human AB of the assay will be low. Falsely high or low GH levels are a result of AB being label- free or bound to non- human AB, e.g. excess GHBP in a RIA or the presence of high- affinity human anti- GH AB would cause inappropriately high GH readings [62]. Regulation of GH Secretion The mechanisms of GH release are complex. The interested reader is referred to reviews of this rapidly evolving field for more details [63 66]. A brief summary of the mechanisms involved will be given here. The major components of GH regulation are growth hormone- releasing hormone (GHRH), which is stimulatory, and growth hormone release- inhibiting hormone (GHRIH), also referred to as somatostatin, which inhibits the pituitary somatotrophs (fig. 1). The site of GHRH production is predominantly in the arcuate nucleus. GHRH binds to a G- protein- coupled receptor at the surface of the somatotrophs. The production of GHRIH occurs predominantly in the paraventricular nucleus of the hypothalamus. Recently, another peptide, ghrelin, GHD Testing 111