2.3.11 Acromegaly

Acromegaly is the condition most often associated with an anterior pituitary tumour, which results from growth hormone and insulin-like growth factor 1 (IGF-1) excess. It causes most characteristically enlargement of the hands and feet (Greek: akron, extremities; megas, great). Gigantism, which is the juvenile counterpart of acromegaly, is also caused by a pituitary tumour secreting growth hormone, but it causes excessive growth before epiphyseal fusion. It occurs less frequently than acromegaly because pituitary tumours in children are much less common than in adults.

Goliath was the first giant to be recorded (290 cm/9 ft 6½ inches). The pharaoh Akhenaten—the iconoclast who moved the capital of Egypt and originated monotheism in favour of the sun—is often suggested to have acromegaloid features, but probably did not have acromegaly. It is more likely that his acromegalic appearances were a family trait and anyway he was fertile. The Irish giant James Byrne, whose skeleton is exhibited in the Royal College of Surgeons of England, was 234 cm. Cushing correctly suggested that he would have an enlarged pituitary fossa. The tallest man recorded was Robert Wadlow, an American who died in 1940 at the age of 22 years (272 cm). Comprehensive historical and illustrated descriptions of acromegaly and gigantism are available (1–3).

Acromegaly was first described in 1886 by Marie (Fig. 2.3.11.1) a pupil of Charcot. Although there had been previous cases described, it was Marie who gave the name to the condition. He did not at the time realize that the pituitary was the cause of the problem and the first recognition of an enlarged pituitary is attributed to Minkowski (1887). The first attempt at surgical treatment was by Caton and Paul in Liverpool (1893). They attempted to relieve the headache simply by surgical removal of part of the skull vault. Harvey Cushing was convinced that acromegaly was a form of hyperpituitarism and he operated for the first time, via the transsphenoidal route, to improve the condition. Radiation therapy was reported first in 1909 by Béclère. The development of radioimmunoassays for growth hormone in the 1960s provided the tools for the more accurate assessment of the disease. Medical therapy with dopamine agonists was introduced by Liuzzi and colleagues in Milan in 1972. In 1986, the first somatostatin analogues were described as providing more effective lowering of growth hormone levels in acromegaly. In 2000 a growth hormone receptor antagonist was shown to be very effective (4).

Several epidemiological studies have been published (5). The mean incidence per million is 3.3 per year with a mean prevalence ranging from 38 to 69 cases per million. More recently, a higher prevalence of about 130 per million has been suggested by a study with more active surveillance for pituitary adenomas (6). Acromegaly occurs in all races with an approximately equal sex incidence. Peak age at diagnosis is 44 but patients with acromegaly can present at all ages. The mean time to diagnosis is 8 years with a range of 6–10 years. Larger, more aggressively behaving tumours secreting growth hormone tend to be present in younger patients. Patients with family history with pituitary adenomas present at an earlier age (7, 8).

Acromegaly is most frequently caused by a pure growth hormone-secreting adenoma. A third of patients with pituitary tumours have mixed growth hormone- and prolactin-secreting adenomas. Very rarely growth hormone and thyroid-stimulating hormone (TSH) are secreted together, causing acromegaly with thyrotoxicosis and a detectable TSH (Box 2.3.11.1).

Less than 1 % of patients with acromegaly have a growth hormone-releasing hormone (GHRH) secreting tumour. This is usually a carcinoid tumour either in the pancreas or in the lung. These are associated with pituitary somatotroph hyperplasia which histologically often gives the clue to the presence of the GHRH-secreting lesion, which may also present on its own accord. In such cases, the pituitary is globally enlarged, with no focal tumour detected. Very rarely hypothalamic GHRH-producing tumours have been described, such as, gangliocytoma. Carcinoma of pituitary secreting growth hormone has been described but is very rare (see below).

Acromegaly can occur as part of a genetic condition due to (1) Carney complex, (2) familial isolated pituitary adenoma (FIPA) (3) multiple endocrine neoplasia type 1 (MEN 1) or (3) McCune–Albright syndrome (see below). Acromegaloidism (insulin-mediated pseudo-acromegaly) refers to the development of acromegaly-like features (e.g. jaw, hand, and feet enlargement) together with acanthosis nigricans caused by very severe insulin resistance. Growth hormone and IGF-1 values are normal (9). Rarely pachydermoperiostosis (OMIM 1671002) or a familial condition with variable acromegaloid features and abnormalities of chromosome 11 (10) can present as differential diagnostic problems.

Somatotroph cells are usually located in the posterolateral region of the pituitary often explaining the cavernous sinus invasion of these adenomas. In normal somatotroph cells the growth hormone-containing vesicles are 400 nm in mean diameter. Somatotroph adenomas can either be sparsely or densely granulated. The sparsely granulated somatotroph adenomas occur more often in young patients, tend to be more aggressive with cells showing less differentiation and have a greater tendency to tumour invasiveness.

Approximately one-third of patients with acromegaly present with hyperprolactinaemia due to increased prolactin secretion from a tumour or alternatively loss of dopamine inhibition from stalk compression because of a macroadenoma. Prolactin secretion from the tumour can be due to mixed somatotroph and lactotroph adenomas, with discrete populations of growth hormone or prolactin-secreting cells or due to mammosomatotroph tumours, which are composed of cells that produce both growth hormone and prolactin. Mixed somatotroph and thyrotroph adenomas are associated rarely with acromegaly and thyrotoxicosis.

There have been at least 10 reported instances of metastasizing pituitary carcinomas secreting growth hormone. Most often the metastases are found in the cerebrospinal axis but they have been described outside the central nervous system. The incidence probably lies between 0.1 and 0.5% of clinically diagnosed anterior pituitary adenomas (11).

The molecular pathogenesis of sporadic growth hormone secreting pituitary tumours is best considered by discussing changes which activate factors leading to increased tumour formation, e.g. oncogenes, or alterations which inactivate cell proliferation controlling genes, e.g. tumour suppressor genes. Amongst the described activating genetic alterations are stimulatory guanine nucleotide-binding protein (G-protein) α-subunit gene (GNAS), cyclin D (CCDN1), fibroblast growth factor receptor 4 (FGFR4), and pituitary tumour transforming gene (PTTG).

The G-protein is involved in the activation of adenylate cyclase which mediates the regulatory actions of GHRH to stimulate growth hormone synthesis and secretion. Missense mutations of GNAS at codons 201 and 227 (termed ‘gsp’ mutations) most commonly result in inhibition of the intrinsic GTPase activity of the α-subunit of the G protein adenyl cyclase which is persistently activated resulting in high intracellular levels of cyclic AMP and its downstream pathway including increased protein kinase A and cyclic AMP-response element binding protein (CREB) activity (12). This results in autonomous growth hormone secretion (Fig. 2.3.11.2). This somatic mutation has been demonstrated in 40% of human growth hormone secreting pituitary adenomas and is the most commonly described the genetic defect. If the GSP mutation occurs in embryonic stage and is found in a mosaic form in various organs contributing to activation of various G_s-coupled receptors, the patient develops McCune–Albright syndrome (see below). Increased PTTG mRNA expression has been demonstrated in somatotroph adenomas and correlates with tumour size. FGFR4 and cyclin D overexpression have been described in pituitary tumours; however, this is not specific for somatotroph adenomas.

Tumour suppressor genes that may be involved in pituitary tumour pathogenesis include the retinoblastoma (Rb) gene, cyclin-dependant kinase inhibitors such as p27 (CDKN1B) and p16 (CDKN2A) as well as growth arrest and DNA damage-inducible protein (GADD45γ) and maternal imprinting gene 3 (MEG3). Some of these proteins are lost in pituitary tumours due to epigenetic mechanism such as hypermethylation. p27 expression is reduced in all types of pituitary adenomas including somatotrophs. GADD45γ is a proapoptotic factor which is lost in growth hormone-secreting adenomas. MEG3 is an imprinted gene encoding a noncoding RNA that suppresses tumour cell growth; it is lost in nonfunctioning pituitary adenomas but not in somatotroph tumours. Aryl hydrocarbon receptor-interacting protein (AIP) germline mutations have been described in families with isolated pituitary adenomas and in vitro studies confirm that loss of function of this protein is in the pathogenesis of these adenomas. Occasionally seemingly sporadic cases are also positive for AIP mutation but the change is detectable in germline DNA and in one of the parents as well (8, 13).

Theoretically it is possible that there is a role of hypothalamic factors and GHRH in the autocrine or paracrine role in growth hormone-secreting tumour pathogenesis, and this has been shown in a mouse model of GHRH overexpression. However, this has not been shown in humans.

This is characterized by polyostotic fibrous dysplasia, hyperpigmented cutaneous patches, and endocrinological abnormalities including precocious puberty, thyrotoxicosis, gigantism, and Cushing’s syndrome. The genetic defect is a somatic mosaicism for the gsp mutation which results in autonomous activation of adenylate cyclase generally causing growth hormone hypersecretion and somatotroph hyperplasia. Growth hormone excess is observed up to 20% of the patients and somatotorph and lactotroph hyperplasia have been described but detectable pituitary adenomas are identified in only few patients (14).

This is an autosomal dominant condition caused by a mutation in the protein kinase A regulatory subunit gene (PRKAR1a on 17q22-24) in 60% of the cases with the other 40% mapped to 2p16. It is characterized by spotty cutaneous pigmentation, cardiac and other myxomas, and endocrine overactivity, particularly Cushing’s syndrome due to nodular adrenal cortical hyperplasia. Similar to McCune–Albright syndrome, abnormal growth hormone dynamics can be detected in a high proportion of cases and somatotroph hyperplasia has been documented but patients only rarely develop true adenomas (14).

This is an autosomal dominant disorder with incomplete penetrance characterised by familial occurrence of pituitary adenomas but no other endocrine abnormality, therefore clearly distinguished from MEN 1 and Carney complex. Most often family members have acromegaly but mixed acromegaly-prolactinoma families and more rarely nonfunctioning adenoma families have also been found. In 30–50% of the cases a mutation can be identified in the AIP gene (15), while in the rest of the families mutations in probably other gene(s) cause the disease. Patients with AIP mutations usually have early-onset disease, the penetrance is 30%, and the responsiveness to somatostatin analogues is poor. In families without AIP mutations the age of onset is higher and the penetrance is lower, with a more mixed picture of the type of adenomas presenting in the family members (6).

This is an autosomal dominant disorder which is described elsewhere (see Part 4). Acromegaly is not the commonest of the pituitary hypersecretory syndromes to occur in MEN 1.

Gross acromegaly is easily recognized. The diagnosis in younger patients is more of a test of clinical acumen. Growth hormone and IGF-1 enlarge everything except the nervous system. The most noticeable feature is usually a change in facial appearance. Vague symptoms such as fatigue may predominate. Increased sweating and sebaceous activity can be noticed in the face. There is enlargement of the supraorbital ridges, prognathism, and macroglossia (see Fig. 2.3.11.4); interdental separation occurs. This together with the obvious changes in the hands and feet often makes the diagnosis easy (see Box 2.3.11.2). Often headache is a typical symptom, more commonly than other types of pituitary adenoma. Patients are commonly recognized by rheumatologists, orthopaedic surgeons (joint pain and abnormalities), dentists (separation of the teeth), neurologists (carpal tunnel syndrome), or by physicians treating the patient’s sleep apnoea, hypertension, or diabetes. Often the symptoms are present and progress insidiously over several years. It can be useful to review serial old photographs to show the presence and progress of subtle facial appearances.

Symptoms of an enlarged pituitary fossa are the same as with nonfunctioning tumours and are discussed on page 209. They include visual filed defects, headache, and pituitary apoplexy (more often in younger patients) can be a rare presenting feature of acromegaly.

The skin on the back of the hand is thickened and this may be a very useful bedside test. Increased sweating occurs in 80% and patients look older than their years.

In the cardiovascular system, hypertension is present in 50% due to a direct effect of growth hormone on sodium absorption, and there is also increased left ventricular muscle mass. Ischaemic heart disease is also present, possibly exacerbated by insulin resistance and is a major cause of morbidity and mortality. Myocardial hypertrophy with fibrosis leading to ventricular dilatation and biventricular failure are features of an acromegalic cardiomyopathy.

Respiratory symptoms are also common and account for part of the increased mortality of the condition. Sleep apnoea may result from significant airway obstruction caused by prognathism, macroglossia, and hypertrophied nasal structures. Difficulty in tracheal intubation is often encountered in acromegalic patients undergoing anaesthesia. There may also be a central element to sleep apnoea and narcolepsy may be a presenting symptom in patients with acromegaly.

In the alimentary tract macroglossia and visceromegaly are common. A high prevalence of colonic polyps in acromegaly is reported and these may progress to colonic neoplasia. Hence vigilance is important and full-length colonoscopy recommended on presentation or at aged 50. With careful preparation and appropriate equipment the technical difficulties due to the enlarged bowel can be overcome. Growth hormone and/or IGF-1 may possess direct mitogenic effects on colonic epithelial cells. The latter is expressed in colonic carcinomas where IGF-1 receptors are present.

The musculoskeletal changes predominantly involve the weightbearing joints. Proliferation of chondrocytes occurs in response to increased growth hormone and IGF-1 levels. The osteoarthritis that subsequently develops can be extremely debilitating and this is one complication of acromegaly that is difficult to reverse.

Increased insulin resistance occurs because of direct anti-insulin effects of growth hormone. Acromegalic patients may develop type II diabetes mellitus and carbohydrate tolerance is considerably improved with successful therapy after lowering of growth hormone. Frank diabetes mellitus occurs in about a third of patients. Hypercalciuria occurs in 80% of patients because of growth hormone being facultative in the synthesis of 1,25-dihydroxyvitamin D. Hyperphosphataemia may occur due to the direct effect of GH/IGF-1 on renal phosphate reabsorption. If hypercalcaemia is detected hyperparathyroidism and MEN 1 (3%) need to be investigated. Multinodular goitre occurs with increased frequency in acromegaly. IGF-1 is a major determinant of thyroid cell growth. Thyroid dysfunction (hyperthyroidism) occurs in acromegaly and is most commonly due to a multinodular goitre but TSH secretion from a mixed pituitary tumour should be considered if the TSH is inappropriately normal/elevated in association with thyrotoxicosis.

Acromegaly is associated with a decreased life expectancy. This was first shown in the 1950s and later it was confirmed that that these patients have an increased cardiovascular and respiratory mortality (16). More recently the possibility of increased mortality due to malignant disease has been raised. Overall mortality of untreated disease is approximately double normal. As the tumours tend to be larger and have a greater frequency for being extrasellar in younger patients, particularly those with extrasellar tumours are more difficult to treat successfully. This applies to all modalities of treatment, including surgery, medical treatments, and radiotherapy.

This increased risk relates to hypertension and diabetes. There is no characteristic lipid disturbance in acromegaly. Before 1966, 50% of acromegalic patients died before the age of 50, cardiovascular disease being the commonest cause of death. Cardiovascular disorders accounted for about 25% of deaths, followed by respiratory (20%) and cerebrovascular disease (15%). More recent data suggest a twofold risk of cardiovascular disease and no increased respiratory mortality (17).

Most previous series show an increased risk of malignant disease (Table 2.3.11.1), but it is interesting that the largest cohort did not show this, although it did show an increased risk of colonic cancer (relative risk 1.68) (18).

The diagnosis of acromegaly is made with observing an elevated IGF-1 level as matched for age and gender, and failure to suppress growth hormone in response to an oral glucose tolerance test (OGTT) usually to a level of less than 1 μg/l (19). But for early detection of the disease when using sensitive growth hormone assays the threshold should be lowered to 0.4 μg/l (20). In patients with acromegaly there may even be a paradoxical rise in growth hormone in response to OGTT. False positives do occur (Box 2.3.11.3) but few conditions apart from adolescence are likely to cause diagnostic confusion. However, in tall adolescents, possibly associated with large growth hormone pulses, growth hormone levels may not become undetectable during an OGTT, thus raising the possibility of acromegaly. In these patients IGF-1 is not elevated.

Growth hormone levels even if elevated, are not individually adequate to diagnose acromegaly. Multiple samples during the day, however, always show detectable levels of growth hormone, whereas in normals 75% of the samples during the day are undetectable. The IGF-1 level is invariably high in acromegaly. Occasionally patients who are very ill with acromegaly and in whom IGF-1 is measured, may not demonstrate an elevation, but this becomes apparent later when they recover from the intercurrent illness.

Insulin-like growth factor binding protein 3 is not so growth hormone dependent as IGF-1 and does not give the same clear differences between acromegaly and normality (Fig. 2.3.11.6). In 80% of the patients with acromegaly there is a paradoxical release of growth hormone (by 50% over basal, or an increment of at least 3 μg/l) after thyrotrophin-releasing hormone (TRH) and less frequently after gonadotropin-releasing hormone (GnRH). This and the paradoxical fall in growth hormone seen in acromegaly in response to dopamine and dopamine agonists are rarely required to confirm the diagnosis of acromegaly.

The suspicion of acromegaly can be based on typical acromegalic features (35%) but often on associated abnormalities such as amenorrhoea, visual field defect, carpal tunnel syndrome, joint problems, or headache. About 50% of patients are diagnosed when seeking medical advice for an unrelated complain.

Growth hormone levels tend to be higher in younger patients presenting with larger tumours (Box 2.3.11.4). In those who present after the age of 50, the tumour is often smaller and intrasellar. There is a relationship between serum IGF levels and the log of the serum growth hormone. Saturation of IGF-1 occurs above a growth hormone level of 20 μg/l whereafter, little further rise in IGF-1 occurs. Plasma GHRH should be measured if an ectopic source of acromegaly is suspected, or if occasionally, pituitary histology reveals hyperplasia.

About 40% of patients present with microadenomas, the rest are macroadenomas that may extend outside the fossa.

Essential hypertension is common in acromegaly, often associated with an increase in intravascular volume and low renin and increased aldosterone secretion. Phaeochromocytoma is not associated with acromegaly; however, it is important to exclude a phaeochromocytoma in a hypertensive patient with acromegaly, particularly prior to surgery.

The ideal treatment will render growth hormone secretion normal, completely ablate the pituitary tumour mass, whilst preserving normal pituitary function resulting in complete reversal of acral and other systematic complications of growth hormone excess. There should be no biochemical or tumour recurrence. No currently available treatment effectively fulfils all these criteria.

Primary treatment is usually surgical. Most often this is accomplished through the transsphenoidal route. If this fails, medical treatments to reduce growth hormone and IGF levels to normal should be initiated (50–60% overall) (Box 2.3.11.5). Usually this is first attempted using an analogue of somatostatin (octreotide or lanreotide). If unsuccessful, cabergoline the best tolerated dopamine agonist is added up to a weekly dose of 3 mg, although one should be aware of possible cardiac valve effects of long-term high-dose cabergoline treatment. Then pegvisomant should be added if possible. At this stage radiotherapy is considered.

Abundant epidemiological evidence suggests that a growth hormone level of 1 μg/l or less is associated with a normal life expectancy (21). The most important determinant of outcome is the most recent growth hormone or IGF-1 level. Normalization of IGF-1 is associated with no difference in survival from a control sample. Other factors which have been associated with increased mortality include duration of symptoms prior to diagnosis, duration of disease, older age at diagnosis, and the presence of cardiovascular disease, diabetes, and hypertension at diagnosis.

After surgery, growth hormone pulses are often not normal. Growth hormone deficiency may occur and in most patients growth hormone secretion is not normal. After radiotherapy too, growth hormone pulses become absent and there is often a constant low-grade level of elevated growth hormone secretion resulting in higher IGF-1 than one would expect from the ambient growth hormone. These facts have led to the concept of a safe growth hormone level (mean of less than 1.7 μg/l) rather than talking specifically about a cure which in terms of normalization of growth hormone secretory dynamics virtually never occurs.

Table 2.3.11.2 shows the effects on growth hormone levels of surgery in various surgical centres throughout the world. It is evident from these figures that the outcome for microadenomas is better than that for macroadenomas. In addition, the criteria used to judge success differ widely. A mean of several growth hormone levels of 1.7 μg/l or less are equivalent to a nadir achieved during oral glucose tolerance of levels less than 0.5 μg/l (22). Given these figures, it is clear that there is quite a wide disparity in outcomes, but the best available figures in the best surgical hands show that between 70% and 90% with microadenomas and between 45% and 50% of macroadenomas should have levels of growth hormone rendered into the safe range with surgery (23).

The most common complication is hypopituitarism. This can involve anterior or posterior pituitary function and complication rates appear to be higher with bigger tumours. New hypopituitarism develops in between 12% and 18% of patients undergoing transsphenoidal surgery for acromegaly. These patients may require lifelong pituitary hormone replacement therapy. Occasionally pituitary function may recover (22). Other complications include transient or permanent diabetes insipidus, cerebrospinal fluid leaks, haemorrhage, and meningitis. Recurrence of acromegaly occasionally occurs (5.5% at 3 years).

Pretreatment growth hormone levels in a large number of series have been shown to affect outcome such that high levels are associated with a less successful surgical outcome. In a series by Sheaves et al. (24) postoperative growth hormone levels fell below 1.7 μg/l in 65% of patients in whom pretreatment growth hormone levels were less than 6 μg/l, and in only 18% of those in whom pretreatment levels were greater than 33 μg/l. Table 2.3.11.2 also shows the effect of tumour size (micro vs macroadenoma) on surgical outcome.

Surgical experience has been shown to have a significant impact on the outcome of surgery. With large numbers of surgeons doing a small number of operations annually, the outcome is less good and in several centres the outcome has been improved considerably following the policy of adopting one or two surgeons to do all pituitary surgery. Complications are also less common with experienced surgeons(23, 25).

Transcranial surgery is occasionally necessary when there is a very large suprasellar extension or a tumour extending out laterally which is unreachable transsphenoidally, although the use of endoscopic surgery increases the reachable areas in the lateral direction. In cases where transcranial surgery is indicated the reduction of growth hormone to safe levels is virtually never obtained.

Octreotide and lanreotide are synthetic octapeptide analogues of somatostatin which share some amino acid homology with it. They exhibit pharmacological effects similar to somatostatin, although with a much longer duration of action than the parent compound (Fig. 2.3.11.7). Unlike the parent compound there is no rebound hypersecretion of growth hormone and other hormone secretions following cessation of their action. There are five somatostatin receptor (SSTRs) subtypes. The main SSTR subtypes on the anterior pituitary are SSTR2 and SSTR5, and octreotide and lanreotide bind specifically with high affinity to these receptors. A newer somatostatin analogue, pasireotide (SOM230), has a wider activity on all SSTRs except SSTR3. Whether newer analogues of somatostatin, like SOM230, which stimulate other somatostatin receptors, are more effective has yet to be established.

Initially somatostatin analogues were given thrice daily, as subcutaneous injections. Longer acting somatostatin analogues have been developed which need to be administered once a month. Octreotide LAR and Lanreotide Autogel are two such analogues. Lanreotide exhibits a two phase pattern with an instant release of the analogue localized at the surface of the copolymer, followed by a second period with a slower and more prolonged liberation by enzymatic breakdown of microcapsules.

The effect of somatostatin analogues on growth hormone production in acromegaly can be predicted by a single subcutaneous dose of octreotide which in responsive patients shows a fall to less than 1.7 μg/l.

Octreotide LAR is started with a dose of 20 mg per month. After 3 months, growth hormone levels are re-assessed and if greater than 1.7 μg/l the dose should be increased to 30 mg, and if less than 1.7 μg/l reduced to 10 mg. Between 50 and 80% of patients on this drug, attain safe growth hormone levels of 1.7 μg/l or less. Around 50% achieve a normal age-related IGF-1. In general, patients starting with high growth hormone levels are less likely to achieve safe values on octreotide or lanreotide than those starting with lower values. Comparison of octreotide LAR and Lanreotide Autogel show similar numbers of patients who attain growth hormone levels of less than 1.7 μg/l (27)Patients with sparsely granulated tumours and patients from families with familial isolated pituitary adenomas are less responsive to somatostatin analogues (8, 28, 29).

Patients on somatostatin analogue therapy can be followed by IGF-1 and by mean growth hormone levels but the response to OGTT is variable (19). For patients who were treated with radiotherapy and are currently on medical therapy 12–24-monthly temporary cessation of medical treatment is suggested for the assessment of growth hormone/IGF-1 status unless it is still high despute on going somatostatin analogues treatment.

Despite suppression of insulin, the effect on growth hormone predominates and in the majority of patients, somatostatin analogues improve carbohydrate tolerance. In contrast to the effect of dopamine agonists, somatostatin analogues usually do not have an effect on prolactin levels. Somatomammotropic tumours treated with long-acting somatostatin analogues may show a fall in prolactin levels as well as growth hormone.

Diarrhoea and abdominal pain occur in 30% of patients to a mild or moderate degree initially but in the vast majority these usually settles (Box 2.3.11.6). The most important chronic side effect is gallstones, which complicates long-term therapy with octreotide and the somatostatin analogues. The rate varies widely between 14% and 60% and probably depends on the length of treatment. They develop because octreotide decreases gallbladder contractility by suppressing cholecystokinin (CCK) release. Bile also becomes abnormal, possibly in relation to prolonged intestinal transit and altered bacterial flora. The abrupt withdrawal of octreotide may be associated with the development of acute pancreatitis or gallstone colic. Otherwise, gall stones developing on somatostatin analogues very rarely cause symptoms (30). Antibody formation occurs but rarely and is very infrequently significant in terms of altering growth hormone levels. Dependency has been described but very rarely. The compound acts at opiate receptors. For this reason, in occasional patients with severe headache it is very effective at relieving this and often at minimum doses. In these patients headache is improved by frequent subcutaneous doses of 100 μg. However, formal studies comparing subcutaneous and long-acting analogues have not been carried out in this context.

Octreotide and lanreotide are currently the best available medical treatments for acromegaly. Most frequently they are used postoperatively if operations have been unsuccessful at rendering growth hormone levels safe. There is increasing interest in the preoperative use of somatostatin analogues either as an alternative to surgery or for a limited time preoperatively with the desire to reduce morbidity and possibly, by shrinking the tumour, improve the surgical cure rate. Prospective studies of octreotide-LAR in treatment-naïve patients with micro- or macroadenomas have demonstrated normalization of growth hormone or IGF-1 levels in 40% to 70% of patients in the first year with rates improving with longer duration of therapy. A reduction in tumour size of at least 20% is seen in 75% of the patients with a significant improvement in signs and symptoms of disease. However, the overall response rates, particularly in patients with small tumours and low growth hormone levels are lower than surgery and their use would need to be prolonged and therefore expensive. They may also be used following radiotherapy, until radiotherapy has effectively reduced growth hormone and IGF-1 levels to normal.

Bromocriptine, cabergoline, and quinagolide are selective agonist at the D₂ dopamine receptors. Their administration results in the paradoxical fall of growth hormone levels in acromegalic patients, while in normals they stimulate growth hormone levels. Bromocriptine is the only dopamine agonist licensed for the treatment of acromegaly but cabergoline is the most potent and best tolerated.

It is not possible to predict the response to bromocriptine or cabergoline. Overall between 10% and 20% of patients have growth hormone levels that are safe on treatment with bromocriptine (usually 20–40 mg daily) or cabergoline (1–3 mg weekly).

Carbohydrate tolerance improves because of the lowering of growth hormone levels and prolactin levels are suppressed to below normal.

Dopamine agonists may cause acute postural hypotension, nausea, and vomiting. Usually these settle with time. Very rarely, particularly on high doses, psychosis and digital vasospasm may also develop. Cardiac fibrosis has been described with the high doses of cabergoline used for Parkinson’s disease and it is recommended that all patients treated with ergot-derived dopamine agonists (eg. cabergoline) have an annual echocardiogram. No effects on the heart have been found in patients with acromegaly and prolactinoma who are routinely given much lower doses.

Dopamine agonists are less expansive than somatostatin analogues and are available orally. When medical treatment is indicated it should theoretically be the case that dopamine agonists are tried first. In practice, because the response rate is low this does not happen. However, it should be noted that occasionally patients who are not responsive to a somatostatin analogue respond to a dopamine agonist.

A novel growth hormone receptor antagonist has been developed and has undergone evaluation for the treatment of acromegaly (4). Reversible binding to the growth hormone receptor leads to inhibition of signal transduction and therefore lowering of IGF-1. This drug is a recombinant protein with structural similarity to wild-type human growth hormone, but substitutions have been made at the sites of interaction with the preformed growth hormone receptor dimer to leave the receptor inactivated and unresponsive to endogenous growth hormone. An arginine to glycine substitution at position 120 is crucial to inhibition of activation of the growth hormone–receptor complex, while pegylation of the protein increasing its half-life from 11 min to greater than 70 h reduces immunogenicity. The drug is administered subcutaneously and since it does not lower circulating growth hormone, serum IGF-1 is the principal biochemical means of monitoring effectiveness of treatment. With daily dosing and adequate dose titration a satisfactory IGF-1 level can be achieved in over 90% of patients (including those who do not respond to somatostatin analogue therapy). The drug is generally well tolerated with no overall significant change in MRI appearances of the pituitary or the development of antibodies to the drug or to growth hormone. Although there has been some concern that pituitary tumour growth can be observed, the majority of patients included in the studies have had pituitary surgery and/or radiotherapy to the exact frequency cannot be assessed but it appears rare. Liver function tests can also become abnormal, so these as well as pituitary size need monitoring on treatment. Lipohypertrophy at the site of injection has been reported. Control of acromegaly is possible in the majority of patients requiring medical treatment with pegvisomant.

Pegvisomant is indicated for patients unresponsive to somatostatin analogues, and the choice is whether to add it to ongoing somatostatin analogue or substitute pegvisomant in place of the somatostatin analogue. The decision depends on individual patient circumstances, e.g. good tumour shrinkage with a somatostatin analogue would be a reason for combination treatment whilst deteriorating glucose tolerance argues for monotherapy (31).

There are two indications for radiotherapy in patients with acromegaly: to control postoperative tumour growth and to control growth hormone secretion and thereby, with time, allow withdrawal of expensive medical treatment. This is particularly the case if drug therapy cannot attain safe levels of growth hormone and IGF-1 (Box 2.3.11.7). In the absence of radiotherapy, medical treatment is possible but by implication, very expensive option.

There is little doubt that external beam radiation (Fig. 2.3.11.8) is effective in lowering growth hormone (32). Recently there has been some controversy, however, when the current criteria for safe growth hormone levels (1.7 μg/l) are applied to the results that are published. Overall growth hormone levels decline exponentially (Fig. 2.3.11.9) from the beginning of treatment. This is a slow process and at 10 years around 50% of patients have a growth hormone level of less than 1.7 μg/l and a normal IGF-1. IGF-1 levels may be slower to normalize than growth hormone, which reflects continuous low-grade growth hormone secretion without pulses.

Various parameters have been suggested which help predict the kind of patients who will respond to irradiation. Although pretreatment concentrations of prolactin do not reliably predict response, the major determinant is the preirradiation growth hormone level. If, in pretreatment, the growth hormone level is 3–10 μg/l it will take a mean of 4.5 years to achieve a growth hormone of 1.7 μg/l, but if the starting growth hormone is greater than 20 μg/l this level will not be achieved for 7 years or more, if ever. IGF-1 levels postradiotherapy are currently not so well studied. Recent data of the UK acromegaly database do suggest that 56% are normal at 10 years.

Radiotherapy, although it does not always control growth hormone and IGF-1 levels, does prevent further tumour growth and it is very rare to observe tumour growth after external beam irradiation in acromegaly.

In general external beam radiotherapy is well tolerated. Hypopituitarism is common and after exclusion of patients with preirradiation hormone deficiency; postradiotherapy, gonadal, adrenal, and thyroid deficiency occur in 50, 35, and 35%, respectively, at 10 years. Pituitary function deterioration develops gradually so that it is necessary for regular (usually annual) assessment. Visual loss and late malignancy are discussed elsewhere and there is no specific increased risk associated with acromegaly. The effect of external pituitary radiotherapy on memory and mental function requires further study.

Gamma knife therapy and SMART (stereotactic multiple arc radiotherapy) radiotherapy are given in a single session to ablate tumours invading the cavernous sinus and are effective methods of delivering radiation therapy to growth hormone secreting tumours (33). There is a theoretical possibility that because of the steep fall-off of irradiation to surrounding tissues, radiosurgery will be less likely to cause second brain tumours and neurocognitive complications. Further studies in this area are needed. The available data suggest that pituitary hypersecretion may resolve faster with gamma knife therapy, but more longer-term data are required both on this and the effect on pituitary function, but significant numbers develop pituitary complications (34).

For macroadenoma, surgery is ‘curative’ in around 55%. Thus surgery renders growth hormone levels safe in this group. Surgery is usually performed for macroadenoma even if surgical cure is unlikely, because debulking surgery improves the outcome of treatment with somatostatin analogues (36). There are a significant number of patients who require further therapy in these circumstances because elevated IGF-1 and growth hormone levels are associated with an increase in mortality and morbidity. In these circumstances somatostatin analogue therapy or dopamine agonist therapy should be considered. Somatostatin analogue therapy is more likely to render growth hormone levels safe, but occasionally in non-somatostatin analogue responsive patients, dopamine agonists may be effective. The usual first lines are octreotide LAR or Lanreotide Autogel or cabergoline, respectively. It is important to consider external beam radiotherapy in patients whose medical therapy does not render growth hormone levels safe. It should also be considered in patients whose growth hormone levels postsurgery are higher than 1.7 μg/l, and in patients with large extrasellar tumours post surgery (see Box 2.3.11.7). In patients with persisting active disease, pegvisomant, either as monotherapy or in combination with a somatostatin analogue, should be considered. Microadenomas are treated similarly but many more are cured surgically.

Patients with acromegaly should be kept under review, either annually or once every 2 years, probably for life. If at surgery, hyperplasia of the pituitary is found and there is no tumour, a GHRH-secreting tumour, most commonly in the pancreas or in the lung needs to be sought or the possibility of McCune–Albright syndrome or Carney complex considered. This is a very uncommon surgical finding, but it is important that the necessary action is taken.

In patients whose growth hormone and IGF-1 levels are rendered safe and normal, annual review is sufficient when growth hormone and IGF-1 are measured. Recurrence is uncommon but may necessitate further surgery, medical or radiotherapy treatment. Once the level of the rest of the pituitary function is established post-surgery, this does not need to be repeatedly tested because it will not change, unless radiotherapy is given.

If surgery is not curative, further medical treatment and radiotherapy need to be considered as above. Monitoring of both growth hormone and IGF-1 levels should take place after radiotherapy, and caution should be taken as after radiotherapy IGF-1 and growth hormone assessments are often discordant (37). After radiotherapy annual assessment of pituitary function should be carried out (see Chapter 2.3.4). Colonoscopy needs to be undertaken on acromegalic patients at presentation or at age 50 but only needs to be repeated if polyps are found or symptoms develop.

There is a general acceptance that the utility of IGF-1 measurement in the diagnosis and monitoring of acromegaly has been compromised by a lack of standardization, in part due to a lack of a recombinant reference material, and inadequate age-related reference ranges. The recent release of a recombinant IGF-1 reference material (WHO 02/254) and the recognition of the need for more robustly established reference ranges offers the prospect of a new generation of IGF-1 assays in which clinicians can have confidence.

In the near term there are likely to be additional pharmaceutical agents. Pasireotide is a novel somatostatin analogue that is at an advanced stage of clinical development while clinical trials have begun on a chimeric molecule directed at the dopamine and somatostatin receptors.

Increasing numbers of patients are recognized with familial acromegaly. Careful biochemical and genetic screening can identify patients at risk and at very subclinical, early stages of the disease where appropriate intervention can prevent morbidity in these usually aggressive adenoma cases.

Surgical techniques have been refined considerably. In the field of radiotherapy, we need a more detailed assessment of the effects of stereotactic radiotherapy, and from the epidemiological point of view, we need to know more about the effects of hypopituitarism on mortality in acromegaly as well as a number of other disease variables, e.g., hypertension, diabetes, and their effect on mortality. The prospect of being able to control virtually every acromegalic patient, in terms of normalization of IGF-1 levels with the new growth hormone receptor antagonists is exciting.