Growth Hormone Releasing Hormone
HISTORY
In 1982, growth hormone releasing hormone (GHRH) became the last of the originally proposed hypophysiotropic factors to be identified structurally. Its existence had been proposed by Reichlin because selective hypothalamic lesions yielded a GH deficiency state and growth failure,3,4 although it was finally isolated from GHRH-producing abdominal tumors rather than from the hypothalamus. In a patient with acromegaly and Turner's syndrome, the acromegaly was due to somatotroph hyperplasia, as it persisted despite transsphenoidal surgery.116 Two different teams isolated a 40 amino acid peptide from a 5 cm tumor in the tail of the patient's pancreas, GHRH(1-40)-OH, then designated GH releasing factor (GH-RF or GRF).117-119 Simultaneously, the Guillemin laboratory sequenced three GHRH peptides from a different tumor: GHRH(1-44)-NH2, GHRH(1-40)-OH, and GHRH(1-37)-OH.120,121 Complete amino acid sequence identity occurred apart from varying C-terminal extensions, indicating possible peptide processing prior to release. The full biological activity resided in residues 1 to 29,122 and by sequence homology, the peptides were members of the glucagon-secretin family. These peptides eventually were demonstrated to fulfill all the requirements of a hypophysiotropic growth hormone releasing hormone.
MOLECULAR AND CELLULAR BIOLOGY
GHRH is produced predominantly by neurons in the arcuate nucleus of the hypothalamus, which send processes to the median eminence, where GHRH is released into the pituitary portal circulation. GHRH then stimulates the pulsatile release of GH from somatotrophs of the anterior pituitary.123 Both GHRH(1-44)-NH2 and GHRH(1-40)-OH can be found in the human hypothalamus124 and in pituitary tumors of acromegalic patients.125 GHRH is also produced in other tissues, where it may serve autocrine or paracrine roles.
GHRH Peptide
GHRH is a member of a family of homologous peptides that in humans includes secretin, glucagon, glucagon-like peptides (GLP-1, GLP-2), vasoactive intestinal peptide (VIP), pituitary adenylate cyclase–activating peptide (PACAP), PACAP-related peptide (PRP), peptide histidine-methionine (PHM, known as PHI in other species where the C-terminal residue is isoleucine), and glucose-dependent insulinotrophic polypeptide (GIP, also called gastric inhibitory peptide).126,127 These peptides are thought to have arisen from a common ancestor through a series of gene duplications127; as the result of sequence and structural similarities, they can interact at each other's receptors to a varying extent.126 The N-terminal (1-29) residues of GHRH are required for receptor binding in the human122 and are 62% identical in the mouse (the most divergent known mammal) and are less conserved in more distant species, including birds, fish, and even protochordate invertebrates.128-130 This is in contrast to related peptides such as PACAP, VIP, and glucagon, which are 100% identical in many mammals and better than 90% identical in more distant vertebrates.127,130 Indeed, because of sequence similarities between PRP and GHRH in non-mammalian vertebrates, some of the peptides from non-mammalian species originally named as GHRH or GHRH-like are now seen to be orthologues of PRP, and not GHRH.131,132 The relative activities of these PRPs at GHRH receptors appear to differ between species131,133; specific receptors that respond preferentially to PRPs have been identified in fish,134 but a corresponding PRP receptor gene appears to be absent in mammals,134 and human PRP has no activity at the human GHRH receptor.
It has been proposed that the active tertiary structure of the GHRH peptide is an amphiphilic α helix that runs from residue 4 onward.135,136 This helical structure, with polar and hydrophobic faces, is presumably stabilized when the peptide is bound to its receptor but is not stable in aqueous solution.137 Circulating GHRH is inactivated rapidly in vivo by dipeptidylaminopeptidase IV (DPP-IV) acting at Ala2,138 and more slowly by oxidation at Met27.122 Medicinal chemists have used the GHRH scaffold to develop peptidic GHRH analogues with increased stability and potency. These efforts have utilized combinations of strategies that include increasing the stability of the α helix with helix-forming residues or ring structures; prolonging the half-life of the peptide in the circulation by introducing unnatural amino acids or polyethylene glycol (PEG) residues to decrease DPP, tryptic, and chymotryptic protease susceptibility; and replacing the oxidizable methionine.122,139-143 An analogue of GHRH has also been developed that couples to endogenous circulating albumin when injected, thus conferring an extended half-life,144,145 and modified GHRH sequences have been expressed in muscle tissue by electrophoration of injected plasmids.146 Substituting the alanine at position 2 with d-arginine was found to produce a GHRH antagonist,147 and subsequently, more stable, higher-affinity versions of this type of antagonist have been developed148 that may prove useful in blocking the mitogenic effects of GHRH.149 Although GHRH acts as a low-affinity agonist at the VIP receptor, GHRH analogues such as N-Ac-Tyr1 and D-Phe2GHRH(1-29)-NH2 have been developed as VIP antagonists.150
Gene and mRNA
Complementary DNA (cDNA) probes have allowed identification of the single-copy GHRH gene on human chromosome 20. Human, rat, and mouse genes span approximately 10 kb of DNA and include five exons. The third exon encodes residues 1 through 31, which are sufficient for the known biological activities of GHRH. However, the human mRNA encodes a 108 amino acid precursor protein, the middle region of which is processed to form the mature GHRH peptide. Brain-, placenta-, and gonad-specific forms of GHRH mRNA have been isolated,151-153 the messages for which are initiated at different gene promotors and result in mRNAs of different sizes, although the encoded precursor protein remains identical. Immunologic evidence shows that the C-terminal fragment of the precursor protein is processed into an additional peptide known as GHRH-related peptide (GHRH-RP), which is expressed in the human hypothalamus,154 where its role is not known, and in rat testis, where it is reported to regulate Sertoli cell function.155 An additional, alternatively spliced mRNA found in rat placenta but not hypothalamus encodes the normal GHRH but includes an altered GHRH-RP.156
Tissue Distribution
In humans and a number of other species, GHRH immunoreactivity is present in the basal hypothalamus, appropriate anatomically for release into the pituitary portal vessels.157 GHRH cell bodies directing processes to the median eminence originate from both the perifornical nucleus158 and the arcuate (rat) or infundibular (human) nucleus.157 GHRH perikarya are also found in the ventromedial nucleus,159 electrical stimlation of which can induce increased release of GH.160 A reciprocal innervation occurs between GHRH and somatostatin neurons in the rat hypothalamus,161 providing the potential for direct communication between the major stimulatory and inhibitory neurons governing GH release. This relationship may participate in the ultradian oscillation of hypothalamic GHRH and somatostatin mRNA.162 GHRH neurons also directly express somatostatin receptors.163 A number of other brain regions outside of the hypothalamus contain immunoreactive GHRH.158,164,165 The ontogeny of GHRH neurons suggests that they appear in the human fetus at between 18 and 29 weeks’ gestation,166 and in the rat on embryonic day 18, reaching adult levels by postnatal day 30.167
Much evidence for GHRH is evident outside the central nervous system in a number of cell types and tissues in humans and in rodents, but its function outside the GH axis remains to be established. mRNA for GHRH, immunoactive GHRH, or bioactive GHRH content is reported in the anterior pituitary, ovaries, testes, placenta, leukocytes, adrenal medulla, pancreas, and gastrointestinal tract,168 and in tumors associated with the GH axis,125,169-171 as well as many other tumor types, including human breast, endometrium, and ovarian.172 Indeed, trace amounts of GHRH mRNA have been found in most rat tissues examined by reverse transcriptase polymerase chain reaction (RT-PCR).173 Studies in the somatotrope found immunoreactive GHRH in secretory granules and in the cell nuclei.174 Additional data demonstrate somatotrope uptake of labeled GHRH into secretory granules, lysosomes, and the nuclear membrane.175
GHRH RECEPTOR
GHRH acts through a high-affinity G protein–coupled receptor (GHRH-R) found on the anterior pituitary somatotroph and coupled to cyclic adenosine monophosphate (cAMP).176 The receptor was cloned from human pituitary tumor and from rat and mouse pituitary,177-179 and it was found to be a member of the G protein–coupled receptor family B, also called the secretin family; the rat and human protein sequences are 82% identical.177 The GHRH-R protein has 47%, 42%, 35%, and 28% sequence identity with receptors for VIP, secretin, calcitonin, and parathyroid hormone, respectively.177,178 The isolated cDNAs encoded a 423 amino acid protein that has seven putative transmembrane domains and a 108 residue extracellular N-terminal domain (after signal peptide cleavage) containing one glycosylation site. The GHRH-R sequence predicts ten extracellular cysteine residues that also are found in the secretin, VIP, and PACAP receptor subgroup; eight of these ten are also conserved in most members of the wider receptor family B. These cysteine residues are proposed to form sulfhydral cross-links that stabilize an extracellular domain involved in hormone binding.180
Cloned pituitary GHRH-R expressed in cell lines demonstrated saturable, high-affinity, GHRH-specific binding and stimulated the accumulation of intracellular cAMP in response to physiologic concentrations of GHRH.177-179 Unlike some related receptors that could signal through both cAMP : PKA and phospholipase C : IP3:PKC pathways, only cAMP activation could be detected, although stimulation of phospholipid turnover was noted in somatotrope cells. A specific GHRH antagonist blocked both binding and second messenger responses.
Data from the cloned receptor were consistent with photoaffinity cross-linking studies of GHRH-R in sheep pituitary membranes that revealed high-affinity binding sites with an apparent molecular weight of 55 kDa and one glycosylation site. After deglycosylation and taking into account the mass of the coupled GHRH analogue, the molecular weight (MW) of the native ovine receptor protein was estimated at 42 kDa,181 in agreement with the prediction from the human cDNA sequence of 45 kDa, assuming cleavage of a signal peptide.178 Further, the binding characteristics of the natural sheep receptor and the cloned human receptor are largely in agreement with a single high-affinity site with a Kd of ≈0.2 nM.
Various radiolabeled forms of GHRH bind to membranes of the pituitary, thymus, and spleen. The dissociation constants estimated in these studies vary wildly from 41 pM182 to 590 nM.183 No binding was measurable in three nonfunctional pituitary adenomas, although consistent GHRH binding to five acromegalic adenomas was seen, with dissociation constants averaging 0.3 nM.184
Studies that attempted to delineate the receptor's GHRH binding domains using chimeric receptor constructs180 or GHRH cross-linking185 suggest that although the large N-terminal extracellular domain plays a major role in GHRH binding, other domains are also essential for ligand selectivity and binding. In vitro studies of a naturally occurring receptor truncation mutant186 and a receptor truncation resulting from alternative splicing187 show a dominant negative effect on GHRH signaling, suggesting that the receptor may function as a dimer .
GHRH Receptor mRNA and Gene
Two GHRH-R mRNA transcripts of approximately 2.5 and 4 kb were identified in rat pituitary, as were 2.0 and 2.1 kb in mouse and 3.5 kb in ovine pituitary.177-179 Further, in mouse, the receptor is expressed in a spatial and temporal pattern that corresponds to expression of GH.179 In the mouse, the first evidence of POU1F1 (a pituitary specific transcription factor, also called Pit-1 or GHF1) expression occurred at embryonic day 14.5, while transcripts encoding the cloned receptor first appeared on embryonic day 16.5.179 Mutations that cause a loss of POU1F1 expression, such as in dw/dw mice, lead to a lack of GHRH-R gene expression and somatotrope hypoplasia.179
The human GHRH-R gene is divided into 13 exons separated by variably sized introns that spread its length to over 15 kb, the complete sequence of which is known.188 Fluorescent in situ hybridization localized the gene to human chromosome 7p14-15.189,190 Studies of the promoter region of the receptor gene found no traditional initiator motifs such as a TATA box.191,192 Putative binding sites for several transcription factors, including POU1F1, Oct-1, Brn-2, NF-1, cAMP response element (CRE), and estrogen receptor response elements (EREs), were identified. An in vitro reporter system demonstrated that expression was enhanced by POU1F1 and glucocorticoids, and was inhibited by estrogen.191 POU1F1 stimulation is consistent with previous studies in Snell and Jackson dwarf mice showing POU1F1 dependence of receptor expression179; five different POU1F1 sites have been shown to contribute to this POU1F1 dependence.193 The glucocorticoid effect on the promotor may be the mechanism for glucocorticoid upregulation of GHRH binding sites194 and receptor mRNA195,196 in rats. The estrogen inhibition of promotor transcription is consistent with observed sexual dimorphism in receptor mRNA expression.197 Studies have suggested that GHRH-R expression is upregulated by GHRH itself198-200; although a putative CRE that could explain this effect was found in the receptor gene promoter, in vitro regulation of the promoter by forskolin could not be demonstrated.191
The structure of the rat GHRH-R gene201 closely matches that of the human but includes an additional exon that would predict an alternatively spliced receptor message that encodes 41 additional amino acids in the third intracellular loop. Indeed, both rat and mouse GHRH-R cDNA clones for this long form have been isolated,177,179 although analysis of rat pituitary mRNA by PCR found evidence of the shorter form only.177 Alternative splicing at the homologous site in the PACAP receptor results in functional receptors that differ in their relative signaling through cAMP and phospholipase C second messenger pathways.202 When the long form of the rat GHRH-R is stably expressed in cell lines, it binds GHRH, but no intracellular signaling could be detected.201
In mice, evidence is seen of an alternative splice variant that encodes a receptor devoid of the first transmembrane domain.179 Alternatively spliced GHRH mRNA encoding a receptor lacking the last two transmembrane domains has been reported in human pituitary tumors and in normal pituitary.203 In the rat, an alternative splice replacing the last five amino acid residues at the C terminus with a new 17-residue sequence has been found by PCR204; this receptor variant appears to signal cAMP normally. No functional role for any of these alternatively spliced GHRH-R messages has been established, although it has been proposed that a truncated receptor variant expressed in tumors can act as a dominant negative in inhibiting GHRH signaling.187,205
Synthetic GHRH antagonists can inhibit the growth and proliferation of a variety of human tumors and tumor cell lines,206-208 which is consistent both with the hypothesis that GHRH can act as a local autocrine/paracrine factor in the stimulation of cell growth209 and with the mitogenic actions of GHRH at the somatotroph.210 The mechanism underlying this action has been unclear because a full-length GHRH receptor could not be detected in the cell lines that responded to GHRH antagonist.208 However, many tumor cells and also normal human prostate express low levels of alternatively spliced forms of GHRH receptor messages not found in the pituitary.208,211 Indeed, a receptor splice variant with an alternative N-terminal domain (SV1) may be the site of action for the antiproliferative effects of GHRH antagonists,212-214 as well as for ligand-independent stimulation of tumor growth.215
GHRH-R mutations resulting in dwarfism have been identified in mouse and humans. The first such mutation was found in the little mouse, a dwarf strain that has an inherited autosomal defect resulting in low levels of GH and pituitary hypoplasia, but is still responsive to exogenous GH. Pituitary cells from these mice would not respond to GHRH but could release GH in response to other activators of cAMP, suggesting a receptor defect.216 After cloning of the GHRH-R cDNA, the mouse gene was localized to the midregion of chromosome 6.217,218 Subsequent sequencing of the receptor demonstrated an Asp-to-Gly point mutation at residue 60 of the receptor's extracellular domain, which is highly conserved in related receptors, mutation of which resulted in complete loss of cAMP signaling. Additional studies demonstrated that this mutant receptor was properly expressed and localized in the cell membrane, but was unable to bind GHRH.219
In humans, a variety of loss-of-function GHRH-R mutations have been identified in patients with GH deficiency and proportionate short stature.220 Two such mutations have been found in large kindreds—one in three distantly related families from India, Pakistan, and Sri Lanka, and another in a Brazilian kindred with more than 100 affected individuals.221 Additional receptor mutations continue to be identified, and it is now suggested that 10% of all human familial isolated GH deficiency type 1 is caused by GHRH-R defects.222 Affected individuals are homozygous for point mutations or deletions in the receptor protein coding region, at an intron splice junction, or in the gene's promoter. In addition, some affected individuals are compound heterozygotes with two different loss-of-function mutations. Individuals heterozygous for an inactivating missense point mutation of the GHRH-R gene did not show a significant reduction in adult stature or in serum IGF-1, but changes in body composition and possibly an increase in insulin sensitivity were reported.223 In contrast to this, two different receptor truncations have been shown to have dominant negative properties in vitro186,187; thus truncation mutations could potentially affect heterozygotes. It is proposed that activating receptor mutations could be associated with acromegaly or adenoma, but screening of pituitary tumors for activating mutations so far has yielded only ambiguous224 or negative results.225,226
GHRH INTRACELLULAR SIGNALING
At the somatotrope, GHRH activates many of the classical signaling systems, including cAMP, calmodulin, calcium mobilization, and phospholipid pathways, indicating a significant commitment of the somatotrope to respond to GHRH. As with many secretory cells, GHRH-accelerated GH release requires both calcium227 and calmodulin.228,229 Intracellular calcium is elevated within seconds of a GHRH stimulus, both in pituitary cells230 and in thymocytes.231 This calcium response is dependent on influx of extracellular calcium rather than on the release of intracellular stores.232 cAMP also signals to the cell nucleus, regulating gene expression via multiple transcription factors.233
cAMP Metabolism
At the somatotrope, GHRH stimulates cAMP accumulation and GH release, and these responses are blocked by somatostatin.234,235 Glucococorticoid pretreatment enhances both the potency and the efficacy of GHRH in driving cAMP accumulation and GH release236; this steroid is necessary for GHRH-induced cAMP accumulation after several days in culture. Adenylate cyclase activity in membranes of normal rat pituitary or human acromegalic tumor237 is enhanced by GHRH in a guanine nucleotide– and calmodulin-sensitive manner.229 Pertussis toxin enhances GHRH-initiated cAMP accumulation and GH release.234 The spontaneous reduction in GHRH-stimulated cAMP levels that occurs over time can be blocked by cycloheximide,234 and the stimulatory ability of GHRH is potentiated by protein kinase C activation.238,239 This indicates that another receptor system that stimulates C kinase may directly enhance the productivity of the GHRH-R–coupling protein-adenylate cyclase complex,240 a candidate for which is the ghrelin (growth hormone secretagogue) receptor.239
Phospholipids
GHRH increases phosphatidylinositol labeling241 and free arachidonate levels242 in the pituitary. Although in most systems, no effect of GHRH on polyphosphoinositide hydrolysis is detectable,243 a report shows that in a specific subclass of porcine somatotropes (low-density somatotropes), GHRH stimulates both cAMP- and inositol phosphate–dependent second messenger pathways.244 Other metabolic pathways involving phospholipid metabolism may be activated by or may modulate GHRH activity.238,245 cAMP metabolism can be dissociated from GH release after GHRH with some phospholipid metabolic enzyme inhibitors, indicating that they may act distal to the cAMP system to evoke exocytosis.238
Mitogenic Signaling
In vivo, insufficient GHRH signaling during development through a GHRH-R defect218 or as the result of GHRH antisera administration246 results in somatotrope hypoplasia. Excess GHRH signaling through tumor expression,116 Gs mutation,247 or transgenic overexpression248 stimulates somatotrope hyperplasia. In vitro, GHRH is a mitogenic signal for somatotrope proliferation.210 The MAPK pathway is a potential mechanism for this action. GHRH dose dependently stimulates tyrosine phosphorylation of MAPK.249-251
GH mRNA and Release Dynamics in Culture
GHRH stimulates the level of GH mRNA,252 the release of newly synthesized GH253 and total GH (stored plus released),254 and the proliferation of somatotropes in vitro.210 The GHRH effect on the somatotrope varies according to the anatomic location of the somatotrope within the pituitary,255 and the GH releasing effect is further enhanced by acute administration of glucocorticoids,256 possibly through increased GHRH binding.194 Like glucocorticoids, triiodothyronine,257 ghrelin,258 and ghrelin mimetics239 can amplify GHRH-stimulated GH secretion. In contrast, IGF-1256 and somatostatin259 are noncompetitive inhibitors of GHRH-accelerated GH release in vitro.
Accelerated GH release occurs immediately after exposure to GHRH234 and remains elevated for the duration of the GHRH pulse,234 albeit at declining rates of release after about 10 minutes.260 This spontaneous decline could occur without GH content depletion260 and could be blocked by cycloheximide, suggesting the participation of a rapidly turning over inhibitory protein.229 The reciprocal interaction of GHRH and somatostatin, as suggested neuroanatomically and by pituitary portal blood measurements,168 results in a greater mass of GH release per GHRH pulse. This has been demonstrated in perifusion culture.261
Picomolar to nanomolar concentrations of GHRH that likely are present in pituitary portal blood262,263 regulate a graded GH response from the somatotrope.118 GHRH also stimulates modest prolactin release at low GHRH concentrations in vitro,264,265 as well as the secretion of a protein known as peptide 23 (identical to pancreatitis-associated protein and a member of the C-type lectin supergene family).266,267 Because GHRH can interact with VIP receptors in intestinal epithelium268 and GH3 cells,269 it is possible that pharmacologic or pathologic levels of GHRH can activate this and other receptor types.
ANIMAL STUDIES
GHRH Release
The pulsatile release of GH is influenced by numerous factors, including nutrition, body composition, metabolism, age, sex steroids, adrenal corticoids, thyroid hormones, and renal and hepatic functions.168 A major common pathway for these factors is seen in their effects on GHRH release from the hypothalamus through direct actions on GHRH neurons, and also through interplay with somatostatin neurons. These effects may be mediated through other factors such as ghrelin, catecholamines, interleukin-1, somatostatin, opioids, leptin, inhibin, and neuropeptide Y (NPY).270,271
GHRH Effects on the GH Axis
GHRH was first demonstrated to stimulate GH release in vivo in anesthetized rats.272,273 These GH responses to GHRH could be enhanced by passive immunization against somatostatin or blocked by passive immunization against rat GHRH.168 It was soon found that GHRH could enhance GH secretion in every vertebrate species tested, including mammals, birds, and fish.
GHRH is necessary for endogenous pulsatile GH secretion, as anti-GHRH antisera treatment eliminated these pulses in rats and sheep.168 However, rhythmic GH secretion persists in an amplitude-miniaturized version in the absence of a GHRH-R signal, at least in men.274 This observation is supported by a gender-specific difference after administration of a GHRH antagonist, with no change in basal secretion in men but a significant decrease in women.275 Antisera to GHRH also decreased statural growth276 and GHRH-R mRNA expression199 in the rat. Conversely, GHRH administered over several days to weeks enhanced body or organ growth and function in experimental animals.277 The effect is particularly striking in mice transgenic for GHRH.278,279 Pulses of GHRH are measured in pituitary portal plasma of unanesthetized sheep with peak values of 25 to 40 pg/mL for a period of 71 minutes.263 Temporal analysis of GH pulses in sheep suggested a complex regulation that can be explained only partially by GHRH and somatostatin pulses and ghrelin.239,280
During development, basal GH responses to exogenous GHRH decrease from postnatal day 1 to day 28 in the rhesus monkey.281 Passive immunization against GHRH shows that endogenous GHRH is an active secretagogue up to day 9.282 In the rat, GHRH injections do not increase GH levels at postnatal day 2,283 whereas stimulatory responses of similar magnitude are measured at postnatal days 10, 30, and 75, as well as at 14 months.284 Likewise, GHRH injections given over 5 days elevate GH biosynthesis in rat pituitaries at postnatal day 10.285 Rat pituitary GHRH-R mRNA expression was highest in early gestation, declined to a nadir at 12 days of age, increased at the onset of sexual maturation, and then declined with aging.286
Significant changes in GHRH status are noted during aging. In the hypothalamus, a reduction in GHRH gene expression and content is seen,287 as is a decrease in GHRH binding to pituitary in 18-month-old rats.288 GHRH-R mRNA is correspondingly decreased in 18-month-old rats but can be brought back toward levels observed in younger animals through treatment with GHRH.289 Decreased GHRH-R expression may contribute to the diminished pituitary response to GHRH in aged male rats290 and humans.291
The GHRH system is also strongly influenced by gender (or gonadal steroids). Hypothalamic GHRH mRNA levels are greater in male than female rats292,293 and are reduced by orchidectomy and increased by testosterone treatment in intact294 or castrated295 male rats. Estradiol has no effect on hypothalamic GHRH mRNA.296 The ability of GHRH to elicit a GH response in vivo varies during the rat estrous cycle297; it is of greater magnitude in the male than in the female284,298 and is strongly sex steroid dependent.298 Somatotropes from male rats likewise have greater cAMP and GH responses to GHRH than do those from female rats when studied in static299 or perifusion298 cultures. Furthermore, the intact and castrate male rat treated with testosterone yields the greatest quantity of GHRH-responsive somatotropes,298 as does direct testosterone treatment of cultured somatotropes.300 Gender differences can be measured at the level of the single somatotrope using the hemolytic plaque assay301; testosterone (administered in vivo) increases secretory capacity and recruits a subpopulation of somatotropes, while estradiol has the opposite effects.302 GHRH receptor message levels are dramatically lower in female than in male rats,168 and estrogen acts at the receptor gene to inhibit GHRH-R mRNA expression.191
In addition to sex steroid effects, free fatty acids303 and GH itself304 reduce the in vivo release of GH in response to GHRH. GH treatment decreases hypothalamic GHRH content in intact rats293 and after hypophysectomy GHRH levels in the hypothalamus rise,305 suggesting feedback regulation of GHRH by GH. Thyroxine replacement can restore hypothalamic immunoreactive GHRH levels reduced by thyroidectomy in rats,168 and triiodothyronine and cortisol can protect against reduced GHRH-stimulated GH release in hypothyroid306 or adrenalectomized rats,307 respectively. Indeed, long-term glucocorticoids in vivo decrease GHRH expression in GHRH neurons of the arcuate nucleus.308
Months of excess GHRH exposure in transgenic mice is associated with increased pituitary mass and mammosomatotrope hyperplasia278 that eventually results in adenoma formation after 12 months.309 This is reminiscent of the clinical findings in patients with ectopic GHRH secretion. What was surprising was the rapidity of this effect; GHRH infusions were capable of inducing enlargement of the anterior pituitary within days in intact, normal rats.310 The dose range of this acute effect has yet to be defined, and this observation does not address the potential risks of replacement of GHRH in deficiency states. In pituitary allograph studies, in orchidectomized hamsters, exogenous GHRH maintains somatotrope size without affecting the percent of GH cells.311
GHRH Effects on Functions Outside of the GH Axis
GHRH stimulates gastrin release and epithelial cell proliferation in the digestive tract.312 It also stimulates insulin, glucagon, and somatostatin secretion from the pancreas.313-315 GHRH enhances non–rapid eye movement sleep in rats.316 GHRH antisera317 or GHRH antagonists318 inhibit sleep, and sleep deprivation enhances hypothalamic GHRH mRNA levels.319 GHRH-Ab treatment of female rats results in osteopenic effects,320 and plasmid-mediated GHRH expression in an animal model of cancer cachexia was able to prevent weight loss.321 Most of these activities, including control of GH status, suggest that GHRH predominantly acts as a nutrient-partitioning hormone, to regulate body composition.
Circulating GHRH in Humans
Following the initial synthesis of GHRH, radioimmunoassays (RIAs) were used to demonstrate that, contrary to hopes that GHRH in the peripheral circulation would be principally of hypothalamic origin and its measurement would thus serve as an index of hypothalamic secretion, most circulating GHRH is not of hypothalamic origin, but instead comes from the gut.322 Further, an RIA ideally would measure intact biologically active hormone. However, as the result of cleavage by DPP-IV, GHRH(1-44)-NH2 has a very short half-life in the circulation of 6.8 minutes, and the metabolite GHRH(3-44)NH2 appears within 1 minute of an IV injection of GHRH(1-44)NH2.323 The biological activity of GHRH(3-44)NH2 is less than 10−3 that of GHRH(1-44)NH2.138 Unfortunately, most RIAs measure GHRH(1-44)-NH2 and GHRH(3-44)-NH2 with equal efficiency, and measurements therefore do not reflect biological activity in the circulation. Most assays are directed to the midportion of the GHRH molecule and therefore do not distinguish between different circulating forms. Besides the RIA, more sensitive enzyme immunoassays for GHRH measurement have been developed.324 One of the few indications for measuring serum GHRH is a GHRH-producing tumor.
GHRH Levels in Acromegaly
Interest in the frequency of ectopic GHRH as a cause of acromegaly has been intense. Two extensive studies have addressed this issue. In a study of 80 patients with acromegaly, 76 had GHRH levels in the normal range325; of the four with elevated levels, one was known to harbor a GHRH-secreting tumor. Extensive evaluation of the other three failed to determine a source for the GHRH. In a second study, 3 of 177 patients with acromegaly exhibited elevated serum levels of GHRH.326 In all cases, GHRH levels were markedly elevated (i.e., in the nanogram per milliliter range), and patients were known to have had previous GHRH-secreting tumors. Thus, although apparently rare, ectopic secretion of GHRH must be considered as a possible cause of acromegaly, and measurement of peripheral GHRH seems prudent as a part of the evaluation. Because it is known that the release of ectopic hormones may be intermittent, and because only 300 pg/mL of GHRH is necessary to stimulate GH release in normal subjects, a single normal or modestly elevated GHRH determination may not exclude ectopic GHRH-associated acromegaly.
The subject of GHRH-producing tumors has been reviewed previously.327,328 GHRH-producing tumors associated with acromegaly are rare. Unique features of patients with acromegaly harboring tumors secreting GHRH included young age, female preponderance, foregut derivation of the tumors, benign biological behavior, small secretory granules in the tumor, and frequent association with multiple endocrine neoplasia type 1 (MEN-1) syndrome. The pancreas and the lung are common primary sites. GHRH-containing tumors that are not associated with acromegaly include those of the gut and thymus, small cell carcinoma of the lung, and medullary carcinoma of the thyroid. Several tumors are plurihormonal. In contrast to somatotroph adenoma as seen in patients with classic acromegaly, the hypophyseal lesion represents somatotroph hyperplasia in acromegalic patients with GHRH-producing tumor. This finding indicates that GHRH not only increases somatotroph secretory activity, but causes somatotroph proliferation. Studies of GHRH-producing tumors are of fundamental importance for obtaining insight into endocrine activity of pituitary somatotrophs and the pathogenesis of GH-secreting pituitary adenomas associated with acromegaly; the importance of GHRH in the origin of acromegaly is still unresolved. Preliminary evidence suggests that the amount of GHRH mRNA expression seen in somatotroph adenomas is associated with the progression and aggressiveness of these tumors.171 The GHRH receptor mRNA is specifically expressed in GH-producing adenomas and somatotrophs.170 To address whether GHRH can produce tumors, transgenic mice expressing the human GHRH gene have been developed. These animals, which had been exposed to excessive quantities of GHRH throughout development and life, developed mammosomatotrope or somatotrope adenomas.329 The significance of these observations in terms of human disease is unclear, and additional studies are needed. Early studies have investigated the beneficial effects of GHRH antagonists in animals and humans. It is interesting to note that in one single reported clinical case, ectopic GHRH secretion was associated with empty sella syndrome.330
Treatment of transgenic mice overexpressing the human GHRH gene with GHRH antagonists resulted in suppression of GH and secretion of IGF-1.331 The relationship between GHRH-secreting tumors and MEN-1 syndrome is controversial; additional studies are required to elucidate whether they represent two distinct entities, or whether GHRH-producing tumors accompanied by acromegaly are only forme fruste manifestations of MEN-1 syndrome. Several cases of acromegaly due to ectopic GHRH secretion associated with MEN-1 syndrome have been described.332-336
Eutopic GHRH Secretion
Occasionally, hypothalamic gangliocytomas may be associated with acromegaly; immunocytochemical staining of such tumors for GHRH has been described. It has been suggested that these tumors should be considered as an unusual cause of acromegaly.337 On occasion, these tumors are intrasellar; in such cases, the observation that somatotrophs are in close anatomic association with neurons suggests that GHRH not only stimulates GH secretion but may also cause adenoma formation. An intrasellar gangliocytoma with somatotroph adenoma has been described, which was strongly positive for gastrin and weakly positive for GHRH. Because gastrin administered intracerebroventricularly increases GH secretion, it has been suggested that gastrin release may act in a paracrine fashion on gangliocytoma to enhance GHRH secretion and thus cause somatotrope adenoma.338
GHRH Levels in GH-Deficient Children
Many reports describing serum and cerebrospinal fluid (CSF) concentrations of GHRH in children with various forms of growth deficiency have appeared. In 22 children with the diagnosis of constitutional short stature (defined as 2 to 3 SD below the predicted mean height for age, peak levels of GH in excess of 10 μg/L during at least one provocation test, and bone age approximating chronological age), basal GHRH levels (8 to 148 pg/mL) were no different from those noted in normal children.339 In addition, in five of nine children, GHRH levels rose twofold 15 minutes after administration of levodopa (500 mg po). In another study of 16 children with idiopathic delayed puberty, the peak serum GHRH concentration following levodopa was 41 ± 10 pg/mL, and this compared with 96 ± 25 pg/mL in children with constitutional short stature.340 Similarly, in patients with hypothalamic hypopituitarism, no increase in circulating GHRH levels was seen after levodopa, a finding that contrasts with responses in normal subjects.341,342 These patients with hypothalamic hypopituitarism do respond to exogenous GHRH administration. Insulin-induced hypoglycemia increased circulating GHRH levels in normal subjects, but not in six patients with isolated GH deficiency.343 In ten children with short stature, GHRH levels increased at 15 minutes after hypoglycemia from 10 ± 0.5 to 17.1 ± 3.1 pg/mL.344 No increase in GHRH was evident after arginine, even though hypoglycemia alone or arginine alone increased GH concentrations.
However, in contrast to children with constitutional short stature, five children with GH deficiency associated with hypothalamic germinomas were reported to have undetectable concentrations of GHRH in the CSF.345 In addition, children with idiopathic GH deficiency have GHRH present in the CSF but at concentrations lower than those in normal children (15.1 ± 1.0 vs. 29.3 ± 2.0 pg/mL; mean ± SEM).
GHRH AS A DIAGNOSTIC AGENT
Until 1985, the use of cadaveric GH was strictly controlled and was limited to use in children with short stature due to severe GH deficiency. Dynamic tests of GH reserve were therefore of clinical significance in pediatric endocrinology and were performed primarily for academic interest in adults with hypothalamic-pituitary disease. With the advent of recombinant human GH, now available in unrestricted quantities, and its approval for use in adults with GH deficiency resulting from hypothalamic-pituitary disease in the United States, Europe, and Australia, the need to diagnose GH deficiency safely and effectively has become an important issue in adult endocrinology.346
Many of the tests used to determine GH status are hazardous under certain circumstances (the insulin tolerance test [ITT] is contraindicated in the elderly and in patients with ischemic heart disease or a history of seizures) or are associated with unpleasant side effects (glucagon causes nausea and delayed hypoglycemia, clonidine is associated with drowsiness and hypotension, and arginine causes dizziness and phlebitis). Tests that are effective in children, such as those using arginine or clonidine as the stimulus, are less effective at releasing GH in adults.347 The ITT is frequently quoted as the gold standard investigation for diagnosing GH deficiency in children and adults, but many endocrinologists shy away from this procedure because of concerns about the effects of hypoglycemia in their patients, and because the test has to be supervised by a physician for its duration.
Growth hormone-releasing hormone (GHRH) and the GH-releasing substances (e.g., growth hormone-releasing peptide 6 (GHRP-6), growth hormone-releasing peptide 1 (GHRP-1), growth hormone-releasing peptide 2 (GHRP-2), hexarelin, MK-0677) are powerful GHSs that are safe and well tolerated in both adults and children. These agents have attracted increasing attention as stimuli used to determine GH status in both adults and children. Typically, tests of GH reserve in both children and adults have utilized procedures that depend on effects mediated by the hypothalamus. For example, agents such as arginine, glucagon, clonidine, or levodopa and insulin-induced hypoglycemia are assumed to elicit a hypothalamic signal that acts upon the somatotrophs to stimulate the release of GH. A lesion of the hypothalamus or the pituitary may produce an abnormal GH response to these stimuli. The use of GHRH to stimulate GH release directly from the somatotrophs allows a theoretical distinction to be made between patients with a pituitary abnormality, who will have an abnormal response to GHRH, and those with hypothalamic lesions, who may respond normally to GHRH. This may be important when the therapeutic strategy for an individual patient is determined.
Administration of GHRH to Normal Individuals
The effects of GHRH and its analogues given as a bolus have been studied in healthy men, women, and children.348-358 Following an IV injection of GHRH, GH levels begin to rise within 5 minutes and reach a peak at between 30 and 60 minutes. The ability of GHRH to release GH is dose dependent,357 the maximal response being observed following a dose of 1 μg/kg or higher.352 In adults, the GH response to GHRH is similar in men and women, although women are more sensitive to GHRH than men; the dose of GHRH required for half-maximal GH secretion is 0.4 μg/kg in men and 0.2 μg/kg in women. The GH response to GHRH in women is not altered by the changes in sex steroid hormones that occur during the menstrual cycle.356
The effect of pubertal development on the GH response to GHRH is slight.359,360 When 68 prepubertal children were compared with 66 children at various stages of puberty, no overall difference in GH response to GHRH was observed.359 In a more detailed study that examined children at each stage of puberty, a slight decrease in GH response to GHRH was seen in boys during midpuberty. Although a similar decrease was not observed in girls, the GH response to GHRH did not differ significantly between the sexes during puberty.360 In prepubertal children who are being evaluated for poor growth, priming the hypothalamic-pituitary-GH axis with sex steroids can normalize a suboptimal GH response to some stimuli. The GH response to GHRH is not affected by priming with estrogen,361 suggesting that sex steroids assert their effects at the hypothalamus, either reducing somatostatin tone or increasing GHRH release—not at the pituitary.
The ability of GHRH to release GH is similar in prepubertal and pubertal children, and in young adults. However, over the course of the adult life span, the magnitude of the GH peak following GHRH declines with increasing age.348,362 The likely mechanism for the age-related decline in response to GHRH is an increase in somatostatin tone.363 Support for this is found in studies that utilize GHRH in combination with pyridostigmine or arginine.364,365 These agents have been used alone as diagnostic tests for GH deficiency, producing a GH pulse by reducing somatostatin tone.366,367 Arginine and pyridostigmine act synergistically with GHRH, producing large GH pulses similar to those seen in younger adults.
The GHRH Test
With any test used in clinical practice, it is important to determine what constitutes a normal response. Accordingly, several studies have addressed this question for the GHRH test. Ranke et al.368 defined the normal range of GH responses to GHRH in 86 children with a normal GH axis as 11.8 to 172.4 μg/L, by determining the mean response ±2 SD. This study also concluded that a GH response of less than 10 μg/L should be used as the diagnostic threshold for GH deficiency in children.
The diagnostic threshold of 10 μg/L has been used in other studies utilizing the GHRH test369-371 and now is generally accepted in pediatric practice. This peak is similar to that used to define GH deficiency when most stimuli are used in children, although it is recognized that different stimuli are not equal in their ability to release GH.359 Such thresholds are defined arbitrarily, despite attempts to rationalize them.372 This may reflect the fact that the diagnosis of GH deficiency in a child depends primarily on the clinical finding of poor growth, and that the stimulation test is used to confirm the presence of GH deficiency.
In adults, the diagnostic threshold used to identify patients with GH deficiency is lower than that used in pediatric practice. For example, a GH peak of less than 3 μg/L during an ITT is considered to be indicative of severe GH deficiency that warrants therapy with exogenous GH. If arginine were to be used as the stimulus, the diagnostic cutoff would be even lower than this.373 In adults, the GHRH test is generally a more potent stimulus of GH secretion, resulting in higher pulses than those seen following ITT, arginine, or glucagon.374 This would suggest that the diagnostic cutoff for severe GH deficiency might be higher than 3 μg/L when the GHRH test is used in adults, but this has not been determined.
The GHRH Test in Children With Short Stature
Normal Hypothalamic-Pituitary-GH Axis.
The GH response elicited by GHRH in children with short stature due to a variety of causes has been well characterized.368,369,375-383 Children who have constitutional short stature with no underlying pathology have a normal GH peak following GHRH administration,377,383 but the timing of that peak may be delayed.378 Children who are short as a result of intrauterine growth retardation also have a normal GH response to GHRH.
GH Deficiency.
Children with GH deficiency defined by the use of conventional tests frequently have a greater response to GHRH than to other tests of GH status.375 In a study of prepubertal children undergoing investigation of abnormal growth, subjects were divided into groups according to their responses to conventional tests. GH status was considered normal if the peak GH response to a conventional test was greater than 10 μg/L; partial and severe GH deficiency were defined as a peak GH response of between 5 and 10 μg/L and less than 5 μg/L, respectively. When a diagnostic cutoff of 10 μg/L is assumed for the GHRH test, 76% of children with partial and 39% with severe GH deficiency had a GH peak greater than 10 μg/L during the GHRH test. Conversely, 10% of children considered to have a normal GH axis had a peak GH response less than 10 μg/L—a figure consistent with the findings of other studies. This study demonstrated a considerable discordance between conventional tests of GH status and the GHRH test. It also indicated that, in a significant proportion of children with GH deficiency diagnosed clinically and confirmed by conventional tests, somatotroph function may be preserved, with the cause of impaired GH status being hypothalamic rather than pituitary dysfunction.
Patients who develop GH deficiency following cranial irradiation also exhibit discordance between the GHRH test and conventional tests of GH status. The GH response to GHRH in such patients was greater than the response to the ITT and the arginine test in 80% of patients in one study.384 This suggests that radiation primarily affects the hypothalamic mechanisms that regulate GH secretion from the anterior pituitary. The magnitude of the GH peak during the GHRH test decreases as time from radiation increases. This may be a direct effect of radiation on the somatotrophs or the indirect effect of long-standing GHRH deficiency, which depletes the available GH pool within the somatotroph. The latter is supported by the observation that priming with GHRH for several days prior to the GHRH test can significantly increase the GH response to GHRH.385
GHRH and the Diagnosis of GH Deficiency in Adults
The few published studies that examined GHRH as a diagnostic test in GH-deficient adults were performed on small numbers of patients, most of whom had childhood-onset GH deficiency. More recently, studies that have investigated GHRH as a diagnostic agent in adults have utilized it in combination with other agents, with the aim of normalizing the GH response across the adult life span (see later).
Limitations of the GHRH Test
The use of GHRH alone as a diagnostic test for GH deficiency is limited by several factors. The discordance observed between the results of the GHRH test and the results of other stimulation tests such as the ITT can lead to difficulties in interpreting the results of the GHRH test. Considerable interindividual and intraindividual variation has been noted in results of the GHRH test. The coefficient of variation for the GHRH test has been reported as 60% for children386 and 45% for adults,374 although the variability diminishes with increasing age in adults.387 The sensitivity of the GHRH test is relatively poor in children with short stature and varies with the severity of the GH deficiency as determined by other tests of GH status. In one study, the sensitivity of the GHRH test was 24% in patients with GH insufficiency (GH peak 5.0 to 10.0 μg/L to conventional tests) and 61% in patients with severe GH deficiency (GH peak <5.0 μg/L).375 The specificity of the GHRH test has been reported to be 85% to 90%.359,375 The ability to interpret the GHRH test is confounded further by the inhibition of the GH response in obese subjects.388,389 This may be particularly important when patients with GH deficiency are assessed, because they have abnormal body composition with a propensity to abdominal obesity. The GH response to GHRH in obese patients with pituitary tumors is reduced, making it difficult to define their GH status accurately, particularly when the GH deficiency is isolated.390
The GHRH test is safe, is associated with few side effects, and requires minimal medical supervision, making it an attractive test for use in the outpatient setting. However, the problems outlined above make it difficult to interpret the results of the GHRH test in clinical practice; therefore, it must be used with caution. A GH peak greater than 10 μg/L does not exclude GH deficiency in a child with poor growth, and further evaluation of the hypothalamic-pituitary axis should be undertaken. A positive result (i.e., a GH peak <10 μg/L) will be indicative of GH deficiency, but because 10% to 15% of normal children fall into this group, further evaluation to confirm the diagnosis of GH deficiency will be necessary. The relatively good specificity of the GHRH test suggests that a positive result is significant, such that the child should be evaluated further with additional investigations and monitoring. At the present time, the primary role of the GHRH test is to determine which children are candidates for GHRH therapy.
Tests Utilizing GHRH in Combination With Other Agents
The observation that GHRH acts synergistically with agents that reduce somatostatin tone has led to the development of tests combining GHRH with arginine, pyridostigmine, or clonidine. GHRH acts synergistically with these agents, producing large GH pulses, the magnitude of which frequently exceeds 50 μg/L in healthy subjects.359,374,391 The addition of these agents to GHRH increases the reproducibility of the test and improves diagnostic accuracy.
GHRH in combination with pyridostigmine or arginine causes profound release of GH. In one study of normal children and adolescents, the normal range for GHRH plus pyridostigmine was 22.6 to 90.0 μg/L (n = 94), and that of GHRH plus arginine, 22.4 to 108 μg/L (n = 81).359 The results of these tests were not influenced by pubertal stage in either girls or boys.359,392 However, a study by Maghnie et al.393 concludes that in patients with acquired childhood-onset GH deficiency, the number of pituitary hormone deficits and patient age affect the GH response to the combined GHRH and arginine test. Additional study suggests that body mass index (BMI) and age do have an impact on the GHRH arginine test in children with idiopathic GH deficiency.394 Both studies were conducted with a small number of patients.
In children with GH deficiency, the GHRH-pyridostigmine test, used with the diagnostic threshold of 20 μg/L, confirmed the diagnosis in 100% of patients with organic GH deficiency (caused by craniopharyngioma) and in 80% of patients with idiopathic GH deficiency.395 GHRH in combination with arginine has been extensively evaluated in adults with pituitary disease in the hope that it will provide a safer alternative to the ITT, which currently is considered to be the gold standard investigation of GH status in adults. In assessing GH secretion, no difference in sensitivity and specificity has been noted in young adults among the Arg + GHRH test, the PD + GHRH test, and the ITT.396 The combination of GHRH with arginine was chosen over pyridostigmine because the side effects with pyridostigmine were unpleasant and the GH response to GHRH plus arginine is not affected by age.397 In adults, the GH response to GHRH and arginine is not affected by gender or age, as similar results are achieved in male and female subjects and in young and old adults.391,397,398 Across the adult life span, the third percentile limit of GH response to GHRH plus arginine is 16.5 μg/L, and the first percentile limit is 9 μg/L.
Adults with GH deficiency all had a peak GH response to GHRH plus arginine that fell below 16.5 μg/L, and 92% of patients had a peak below 9 μg/L. GH peaks achieved during the GHRH-arginine test correlate well with those obtained during the ITT, although the absolute GH response to GHRH plus arginine is considerably greater than that to the ITT. Of seven patients who had achieved a peak GH greater than 9 μg/L during the GHRH-arginine test, six had achieved a GH peak greater than the diagnostic threshold for the ITT (3 μg/L). Thus, the authors proposed that the diagnostic cutoff for the GHRH-arginine test should be 9 μg/L for severe GH deficiency in adults and 16.5 μg/L for GH insufficiency. Comparison of the sensitivity and specificity of six different tests—ITT, arginine (ARG), levodopa (L-DOPA), ARG + L-DOPA, ARG + GHRH, and IGF-I measurement—in the diagnosis of GH deficiency found the greatest diagnostic accuracy with the ARG + GHRH test and the ITT. The former was preferred over the latter because it produced fewer side effects.399 The GRS consensus of 2007400 suggested that GHRH + Arg, GHRH + Ghrelin, and glucagon tests are as sensitive as the ITT in GHD, provided that appropriate cutoff limits are considered. Clinical practice guidelines of the Endocrine Society401 suggest that ITT and the GHRH-arginine tests have sufficient sensitivity and specificity to establish the diagnosis of GHD. In the clinical situation of radiation-induced GH deficiency, ITT should be used instead of the GHRH-arginine test.402 GHRH directly stimulates the pituitary; therefore it can give a falsely normal GH response in patients with GHD of hypothalamic origin (e.g., those having received irradiation of the hypothalamic-pituitary region). In patients with new-onset GHD (within 10 years) (e.g., that due to irradiation), testing with GHRH-arginine may be misleading. The guidelines also suggest that in the presence of deficiencies of three or more pituitary axes, the occurrence of GHD is very likely, and provocative testing is optional. The validated cutoff level for GHD in adults for the ITT and the glucagon test is a peak GH response <3 μg/L; this has not been validated in obesity. For the GHRH + Arg test, the following cutoff levels have been validated, depending on the BMI: <25 kg/m2, peak GH <11 μg/L; 25 to 30 kg/m2, peak GH <8 μg/L; >30 kg/m2, peak GH <4 μg/L.403 The cutoff level validated for the GHRH + GHRP-6 is 10 μg/L in lean patients and 5.0 μg/L in obese patients.404,405
GHRH THERAPY IN GH-DEFICIENT CHILDREN
The potential of GHRH as a therapeutic agent for GH-deficient children has been examined in several studies. Most GH-deficient children with short stature and growth failure have a disorder of hypothalamic regulation of the pituitary rather than a defect of the somatotroph. In these children, injections of GHRH, as well as other GHSs, might be a useful treatment option. One of the first studies published in 1985 reported the use of GHRH (1-40)-OH administered sc with a peristaltic pump every 3 hours to two children with organic hypopituitarism (post-traumatic, hydrocephalus).406 Children received 3-hourly doses of 1 or 3 μg/kg GHRH for 6 months. Both children increased their growth rate by 1.5- to 6.0-fold compared with the growth rate before administration of GHRH(1-40)-OH. The rationale for using this dose regimen was based on the observation that, in growing children, five to nine pulses of GH are detected every 24 hours.
Since that time, several studies have been performed to evaluate the benefits of GHRH therapy in GH-deficient children with different GHRH injection regimens and different doses.385,407-423 However, the groups of children who were treated were not homogeneous, and their diagnoses varied from GH deficiency and short stature to normal variant short stature without GH deficiency. The GHRH preparations used included GHRH(1-40)-OH, GHRH(1-44)-NH2, and GHRH(1-29)-NH2.
GHRH Given by Pump
Response to the administration of GH with a pump varied from 71% to 100%. The growth rate on GHRH therapy varied from 6.2 to 10 cm/yr over the first 6 months and was maintained for up to 5 years. Growth velocity appears to be related to the total daily dosage, which ranged from 10 to 2150 μg/day. So far, no studies have evaluated whether it is the frequency of administration or the total daily dose that has the greatest impact on therapeutic outcome. The first study that looked at administration by pump or single injections of GHRH was published in 1988.407 It described the effects of different routes of GHRH administration in 24 GH-deficient children. GHRH (1-40) was given either by pump every 3 hours sc or every 3 hours sc overnight only. Alternatively, GHRH was given twice daily at a dose of 1 to 4 μg/kg per dose for 6 months. In all three circumstances, the growth velocity increased between 1.8- and 2.9-fold, with greatest effects seen during pump therapy with injections given every 3 hours.
Unfortunately, no long-term comparative studies have examined the growth-promoting effects of GHRH with GH. Two short-term studies (6 months’ treatment) provided conflicting results: One suggested a comparable growth response with GHRH similar to that with GH treatment,408 whereas the other suggested that GHRH was less effective.409
GHRH Given sc Twice a Day407,410-417
The most impressive results with GHRH injections given twice daily were achieved in a multicenter study using GHRH(1-44)-NH2 in 20 GH-deficient children.414 All children responded with accelerated growth velocity, which increased from 3.6 cm/yr before treatment to 8.6 cm after 6 months and decreased to 8.1 cm after 12 months. Ogilvy-Stuart et al.417 reported a significant beneficial effect of GHRH(1-29) given at a dose of 30 μg/kg/day over 1 year, which resulted in a 1.8-fold increase in growth velocity per year in nine children with radiation-induced GH deficiency.
The effects of GHRH (1-29) therapy in growth retardation caused by chronic renal failure were examined in nine children by Pasqualini and coworkers.416 Children were treated conservatively or with dialysis, or underwent renal transplantation, and GHRH(1-29) 52 μg/kg/day was administered for 3 to 6 months. Growth velocity increased from 3.8 to 8 cm/yr.
GHRH Given sc Once a Day
Several groups have investigated the effects of GHRH given sc by once-daily injection.385,418-423 A study investigating 110 GH-deficient children treated with a dose of 30 μg/kg/day of GHRH(1-29)423 showed a significant increase in linear growth velocity, with 4.1 cm/yr measured before treatment and 7.2 cm/yr after 1 year of treatment (Fig. 23-9). The largest number of children was investigated in the GHRH European Multicenter Study,420 which reported on the treatment of 111 GH-deficient children. Height velocity increased from 3.8 cm/yr before to 6 cm/yr during 6 months of treatment. The group of Bozzola419 reported an increase in growth velocity from 3.5 cm/yr to 7.3 cm/yr after 6 months of treatment in 10 GH-deficient children with GHRH(1-44). Optimal growth velocity was observed during the first 9 months of therapy in most studies.418 Lanes and Carrillo422 reported that therapy with GHRH(1-29) in 16 prepubertal GH-deficient children given once daily sc over 12 to 24 months resulted in a significant increase in growth velocity.
Which Children Should be Considered for GHRH Therapy?
The use of GHRH(1-29)-NH2 in the treatment of idiopathic GH deficiency in children with growth failure was approved by the U.S. Food and Drug Administration (FDA), but due to lack of relative efficacy compared to pharmalogical doses of growth hormone it was withdrawn and is not currently being used. The development of long acting preparations may reverse this.
Adverse Effects
Development of antibodies during GHRH treatment has been described.407,420,423 Treatment-related adverse events, including local injection reactions characterized by pain, swelling, or redness, and headache, flushing, dysphagia, dizziness, hyperactivity, and urticaria, have been reported. Few data on thyroid function have been reported in children with GH deficiency. One study reported an increase in thyroid replacement requirements in one of seven patients424; another study reported a 5% incidence of hypothyroidism.407 The mechanism underlying the change in thyroid function is unclear, but it has been hypothesized that an increase in GH levels results in an increase in somatostatin tone with subsequent inhibition of TSH secretion.425
GHRH TREATMENT IN ADULTS
Optimal GHRH therapy requires the functional integrity of somatotroph cells.426 Although in GH-deficient adults this is no longer the case, older adults in whom somatotroph hyposecretion is thought to be caused by decreased activity of GHRH-secreting neurons427 still have an intact hypothalamic-pituitary axis. In addition, the pituitary GH-releasable pool in the elderly is comparable with that in young adults.
In fact, the few available studies dealing with the administration of GHRH in adults have been performed in the elderly to investigate whether GHRH treatment could counteract the age-dependent decline in GH. Twice-daily injections of GHRH(1-29) for 2 weeks,428 as well as continuous GHRH(1-44) infusions for 2 weeks, in healthy older men429 partially reversed the age-related decrease in GH, as well as IGF-1 levels. Another study, performed over 6 weeks,430 suggested that administration of GHRH to the elderly might attenuate some effects of aging on muscle strength.
The half-life of GHRH after iv injection is about 10 to 12 minutes, which is one of the limitations of the use of GHRH as a therapeutic. The long-acting analogue of GHRH, CJC-1295, contains full biological activity of GHRH(1-44)NH2. It binds to endogenous serum albumin after sc administration, which prolongs its duration of action, and it is modified by substitution of four amino acids, which makes the compound resistant to proteolytic cleavage. Subcutaneous injection of CJC-1293, another GHRH analogue, in healthy adults resulted in increased GH and IGF-1 levels 1 week after injection, which was explained mainly by the increase in trough levels.431 Another study432 showed that in healthy volunteers aged 21 to 61 years, elevated GH and IGF-1 serum concentrations were present for 6 and 14 days, and that tachyphylaxis did not occur with this compound. The sc administration of another compound, PEG-GHRH, increased GH levels in healthy adults, an effect that was present for 12 hours.433 Overall, additional studies are necessary to assess the safety profile of these compounds and its short-term and long-term effects on body composition.