Clinical Presentation and Diagnosis: Growth Hormone Deficiency in Adults

In the United States, growth hormonedeficiency (GHD) affects 50 000 adults,with 6000 new cases yearly.¹ Patientswith GHD have decreased or absent growthhormone (GH) production as a result ofhypothalamic or pituitary disorders resultingin underactive pituitary gland function(ie, hypopituitarism). GHD is distinct fromsomatopause, a term that describes the gradualdecline in GH production through normaladulthood. Adults with hypopituitarismroutinely receive replacement cortisol, thyroidhormone, and gonadal hormonereplacement therapy. Until the past decade,GH replacement therapy had been primarilyreserved for pediatric use. GHD in adults,however, represents a serious clinical disorder,which is distinct from pediatric GHDand can be treated with recombinant humanGH replacement therapy.

GH is produced in the pituitary gland,which is located at the base of the brainbehind the sphenoid sinus in a small bonycavity called the sella turcica (Figure 1). Thepituitary gland secretes hormones from 2distinct lobes, each derived from differentembryonic tissue (Figure 2). These hormonesaffect numerous body systems.

The posterior lobe of the pituitary gland,an extension of the hypothalamus, secretes2 hormones: arginine vasopressin (alsocalled antidiuretic hormone), which stimulatesthe kidney to reduce urine output, andoxytocin, which causes uterine contractions.The anterior lobe secretes 6 hormones:luteinizing hormone, whichstimulates the secretion of sex steroids fromthe gonads; follicle-stimulating hormone,which stimulates ovulation and sperm production;prolactin (PRL), which targets themammary glands to stimulate milk production;adrenocorticotropic hormone, whichtargets the adrenal cortex to cause glucocorticosteroidproduction; thyrotropin influencesthe production of thyroid hormones;thyroid stimulating hormone (TSH) targetsthe thyroid; and, finally, GH (also calledsomatotropin), which targets many tissues topromote growth and control protein, lipid,and carbohydrate metabolism. Of all theanterior pituitary hormones, only GH andPRL act independently and not through a targetendocrine gland. The effects of an excessor deficiency in hormone production are listedin Table 1. The typical managed careorganization formulary includes drug therapiesfor all of the hormone deficiency states,and for several of the hormone excess states.

The pituitary is the master gland, but it isalso controlled by hormones originating inthe hypothalamus, an area of the brain justabove the pituitary. Anatomically linked tothe pituitary through a funnel-shaped structurecalled the infundibulum, the hypothalamuscontrols pituitary function by secretionof releasing and inhibiting factors. GH releaseis stimulated by GH-releasing hormone(GHRH) and inhibited by somatostatin.During a 24-hour period, pulses of GHRHand somatostatin stimulate or inhibit thepituitary to release GH in discrete bursts,with a distinct diurnal phase resulting inmost of the 24-hour GH secretion duringsleep. Most healthy individuals, therefore,have little measurable GH secreted duringmuch of the daylight hours. Measurement ofa single blood sample for GH is therefore nothelpful in making the diagnosis of GH excessor deficiency. When GH deficiency is considered,stimulation testing (sometimestermed "provocative testing" or "dynamichormone testing") is usually performed.When GH enters the circulatory system itattaches to GH receptors in virtually all bodytissue to produce and stimulate local insulin-likegrowth factor 1 (IGF-1) production. Atthe tissue level, many of the effects of GHare mediated by IGF-1. In addition, regulationof GH production by the pituitary iscontrolled by negative feedback from IGF-1.

The level of GH production normallyvaries among individuals and groups ofindividuals. For example, premenopausalwomen produce more GH, and, as peopleage, production of GH gradually declines.Further, the extent of excessive visceral fatis negatively correlated with GH production,whereas exercise appears to be positivelycorrelated.² There are many different etiologiesfor GHD in adults. Table 2 describes thecauses of pituitary or hypothalamic damagethat results in a decrease in GH secretion.The majority of patients with pituitary hormoneabnormalities present initially with apituitary tumor or other sellar lesions.Pituitary tumors may secrete an excess of 1pituitary hormone or be clinically nonfunctioning,producing clinical problems by masseffect, such as headaches or visual fieldabnormalities accused by compression ofthe optic chiasm. Pituitary hormone deficienciesmay be secondary to compressionof the normal gland by a cyst or tumor, orfrom treatment of the tumor, which mayinclude surgery and radiation. Hypopituitarismmay also result from other causes,such as head trauma, infiltrative hypothalamicdisorders, or infection. In addition, forsome children with GHD, the disorder persistsin adulthood.

In adults, GHD can produce metabolic disturbancesthat may compromise the patient'shealth and quality of life and increase cardiovascular(CV) risk. The most salient featuresof GHD in adults include decreased lean bodymass, increased visceral fat and subcutaneousfat, decreased bone mass, and hyperlipidemia.² GHD has been linked to a higher riskof bone fractures, an increase in carotidartery intimal thickness, and elevations incertain markers of CV risk, among them Creactiveprotein and homocysteine. It hasbeen shown that the degree of elevation inlipid level³ and the severity of bone mineralloss correlate with the severity of GHD.⁴Epidemiologic studies have clearly shownthat adults with hypopituitarism, includingGHD, display an increased risk for CV andcerebrovascular disease and premature mortality(Figure 3).^5,6

In the transitional patient, GHD caninduce deleterious metabolic events similarto those in adults. A 2-year, placebo-controlledtrial examined 64 young adults(mean age, 23 years) who had pediatricGHD, current GH levels of less than 5 μg/L,and were not taking GH for an average of 5.6years.⁷ At baseline, 22% of these patientsdemonstrated evidence of below averagebone mineral density, 59% were overweightor obese, and 45% had total cholesterol levelsof more than 200 mg/dL.

For the clinician, diagnosing GHD can bea daunting process. Because GH is secretedin a diurnal pulsatile manner and has a shorthalf-life of only 19 minutes, it is frequentlyundetectable in blood samples withoutprovocative testing. Numerous pharmacologicagents can be used to assess GH productionand secretion by the pituitary inadults (Table 3). These include insulin, arginine,levodopa (L-dopa), arginine plus L-dopa,arginine plus GHRH, and the glucagontest. None display perfect sensitivity andspecificity; however, the insulin tolerancetest (ITT) and arginine-GHRH are excellenttests. A cut point of <5 &#956;g/L has been suggestedto optimize sensitivity of detection ofGHD without sacrificing too much specificity.² The choice of diagnostic test by theendocrinologist depends on many factorsand relates to the clinical setting.

The ITT is considered the gold standardfor determining GHD.² Insulin-inducedhypoglycemia provokes the pituitary tosecrete GH. Once basal glucose levels arelowered, GH levels are assessed at 30, 45,and 60 minutes after insulin administration.However, with ITT, the plasma glucose levelsuggested as an optimal GH stimulus, <40mg/dL, is a level well below normal (70-120mg/dL) and may produce symptoms andsigns of hypoglycemia and some serious sideeffects such as seizures. The ITT requiresstable and adequate hormonal replacementfor other hormone deficiencies and the presenceof obesity may result in false positives.The ITT is also contraindicated in patientsolder than 55 years of age and in patientswith an abnormal electrocardiogram, a historyof ischemic heart disease, or seizuredisorder. Because the ITT is technicallydemanding, the physician's personal experiencewith this test is critical to its success asa diagnostic tool.^8,9

Alternative GH provocation tests includethe use of GH stimulatory substances, suchas arginine, a somatostatin antagonist, withor without GHRH, and L-dopa, alone or incombination with arginine. The combinationof arginine plus GHRH is considered anexcellent alternative to ITT for mostpatients.^8,10 ITT produces a global centralnervous system challenge based on acuteglucose deprivation; in contrast, the arginine-GHRH test is a more direct challenge tothe pituitary and does not involve a comprehensiveneurologic response. Because ofthese differences and the fact that performingan ITT is a higher risk procedure andrequires more personnel, the arginine-GHRH test is likely the overall test of choicein clinical practice. In a study that included157 patients with GHD, a GH response of 9&#956;g/L or less with an arginine-GHRH test wasdiagnostic for severe GHD and correlatedwith lipid abnormalities.⁵ Another study hassuggested a cutoff of 4.1 &#956;g/L for confirmingthe diagnosis of GHD.⁸ It is important tonote that the arginine-GHRH test may befalsely normal in patients with hypothalamicdisease, such as caused by radiation orinfiltrative disorders. Compared with ITTand arginine-GHRH, arginine alone and Ldopaalone or in combination with arginine,as well as clonidine alone, are less specificfor making the diagnosis of GHD unlessvery low cut points are used.⁸ Clonidinealso results in very poor diagnostic performance.Hypotension is a concern whenperforming the clonidine and L-dopa test.The glucagon stimulation test is lengthy,requiring multiple blood samples, and is lesssensitive than ITT.

The dynamic test using L-dopa is of academicinterest because it is not available forclinical use and it poorly diagnoses adultGHD. We should, therefore, consider thisstimulation test as a very crude screening testfor GHD. The stimulation test using clonidinehas evidence that it is extremely poor in discriminatingfor GHD in adults.¹¹

IGF-1 levels are a valuable biochemicalmarker of GH secretion. Low IGF-1 levels inthe setting of hypothalamic-pituitary diseasesuggest GHD. Low IGF-1 may also result fromcauses other than GHD, such as malnutrition,diabetes, hypothyroidism, or liver disease.However, IGF-1 levels are a poor diagnostictest for adult GHD if in the normal range.⁹ Anormal IGF-1 level does not exclude GHD,and, nonetheless, once GHD is documented,IGF-1 levels are useful in following treatmentand evaluating GH dose adjustments.

The American Association of Endocrinologists,the Growth Hormone ResearchSociety, and the British Endocrine Societyhave proposed guidelines for the diagnosis ofadult GHD. These guidelines (Table 4) allrecommend the ITT as the test of choice butvary somewhat based on variations in practiceamong members of these societies.

Difficulty in conducting GH stimulationtesting is often a barrier to diagnosing GHD.Conducting either an ITT or arginine-GHRHstimulation test with approximately 6 bloodsamples during a 2-hour period requiresadded, expensive personnel costs, especiallyfor those in solo practice. The majority ofclinicians in contact with patients who haveGHD do not have residents, fellows, or dedicatednurses to perform these intensive testingprotocols. In addition, the specialequipment needed for these tests, such asinfusion pumps, intravenous solutions, andexam rooms, may not be readily available.Therefore, some patients who are at risk forGHD do not get treated because they do notundergo diagnostic testing. A recent pilotstudy provided a regional endocrine testingfacility for practicing endocrinologists basedon the premise that patients belong to thereferring endocrinologist, all data goes backto the patient through the referring physician,and no patient recruitment occurs withthe testing. This pilot has been successful inhelping endocrinologists gain access to highqualitytesting facilities.¹²

When GHD is highly probable, such as inpatients with a history of pituitary disease,the presence of 3 or 4 other pituitary hormonedeficiencies, and low IGF-1 levels,stimulation testing may not be necessary.Indeed, the probability of GHD in adultsranges from 91% to 100% in the presence of3 to 4 other pituitary hormone deficiencies(Figure 4).¹¹

Because GHD in adults is a deficiencyassociated with known morbidity, higher CVmortality, and there is an approved replacementmedication available, it is importantthat diagnosis and treatment be offered toappropriate patients. The possibility of GHDshould be considered in patients with otherpituitary hormone deficiencies and thosewho have had surgery in the sellar area. Inaddition, patients with pituitary-hypothalamicradiation exposure, including brain irradiation,childhood-onset GHD, or headtrauma are prime candidates for testing aswell. It is controversial whether testing forGHD is appropriate in patients with systemicsymptoms or signs consistent withGHD (ie, low energy, unexplained metabolicsyndrome, or display inexplicably low bonedensity in the absence of other causes).Although more than 90% of adults with GHDhave some form of pituitary-hypothalamicdisease, the remaining 10% are idiopathicwith normal pituitary magnetic resonanceimaging findings. Once identified, GH treatmentshould be weighed against alternativeoptions, such as treatment with statins andbisphosphonates.¹² Further research willdetermine the importance of evaluating thispopulation. Currently, it is important toidentify and treat patients with pituitaryhypothalamicdisease and a high likelihoodof definite GHD.

Endocr Pract

1. Gharib H, Cook DM, Saenger PH, et al. AmericanAssociation of Clinical Endocrinologists medical guidelinesfor clinical practice for growth hormone use inadults and children—2003 update. . 2003;9:64-76.

J Clin Endocrinol Metab

2. Bengtsson BA, Eden S, Lonn L, et al. Treatment ofadults with growth hormone (GH) deficiency with recombinanthuman GH. . 1993;76:309-317.

J Clin Endocrinol Metab.

3. Colao A, Cerbone G, Pivonello R, et al. The growthhormone (GH) response to the arginine plus GH-releasinghormone test is correlated to the severity of lipid profileabnormalities in adult patients with GH deficiency. 1999;84:1277-1282.

J ClinEndocrinol Metab.

4. Colao A, Di Somma C, Pivonello R, et al. Bone lossis correlated to the severity of growth hormone deficiencyin adult patients with hypopituitarism. 1999;84:1919-1924.

Clin Endocrinol (Oxf.)

5. Bulow B, Hagmar L, Mikoczy Z, Nordstrom CH,Erfurth EM, et al. Increased cerebrovascular mortality inpatients with hypopituitarism. 1997;46:75-81.

Lancet.

6. Rosen T, Bengtsson BA. Premature mortality due tocardiovascular disease in hypopituitarism. 1990;336:285-288.

J Clin Endocrinol Metab.

7. Underwood LE, Attie KM, Baptista J, GenentechCollaborative Study Group. Growth hormone (GH)dose-response in young adults with childhood-onset GHdeficiency: a two-year, multicenter, multi-dose, placebo-controlledstudy. 2003;88:5273-5280.

J Clin Endocrinol Metab.

8. Biller BM, Samuels MH, Zagar A, et al. Sensitivityand specificity of six tests for the diagnosis of adultGH deficiency. 2002;87:2067-2079.

Lancet.

9. Hoffman DM, O'Sullivan AJ, Baxter RC, Ho KK.Diagnosis of growth-hormone deficiency in adults. 1994;343:1064-1068.

J ClinEndocrinol Metab

10. Consensus guidelines for the diagnosis and treatmentof adults with growth hormone deficiency: summarystatement of the Growth Hormone Research SocietyWorkshop on Adult Growth Hormone Deficiency. . 1998;83:379-381.

J Clin EndocrinolMetab

11. Hartman ML, Crowe BJ, Biller BM, et al. Whichpatients do not require a GH stimulation test for thediagnosis of adult GH deficiency? . 2002;87:477-485.

Ann Intern Med

12. Isley WL. Growth hormone therapy for adults: notready for prime time? . 2002;137:190-196.

Download PDF: Clinical Presentation and Diagnosis: Growth Hormone Deficiency in Adults