Diagnosis and Treatment of Pituitary Tumors

Pamela U. Freda and Sharon L. Wardlaw

Department of Medicine, Columbia University College of Physicians and Surgeons, New York, New York 10032

Address correspondence and requests for reprints to: Sharon L. Wardlaw, M.D., Department of Medicine, Columbia University College of Physicians and Surgeons, 630 West 168th Street, New York, New York 10032. E-mail: sw22{at}columbia.edu


    Introduction
 Top
 Introduction
 PRL-Secreting Pituitary Tumors
 GH-Secreting Pituitary Tumors
 ACTH-Secreting Pituitary Tumors
 Glycoprotein Hormone (TSH, LH,...
 Summary
 References
 
THE DIAGNOSIS and treatment of pituitary tumors has been improved by a number of recent advances. Diagnosis of hormone hypersecretion has been facilitated by the development of more sensitive and reliable assays, especially for ACTH and insulin-like growth factor I (IGF)-I. Pituitary tumors can now be visualized more accurately due to continued improvements in magnetic resonance imaging (MRI) techniques with gadolinium enhancement. Petrosal sinus sampling can be used to measure hormone levels in the venous blood draining the pituitary. This has proved helpful in the differential diagnosis of Cushing’s disease. Treatment of pituitary tumors has been improved by advances in transsphenoidal surgery and radiotherapy and by the development of remarkably effective drugs for PRL and GH-secreting tumors. The primary therapeutic objectives are to normalize levels of hypersecreted hormones and to reduce tumor size to prevent damage to normal pituitary tissue and surrounding parasellar structures, especially the optic chiasm. This short review describes our current approach to the diagnosis and treatment of hormone- and nonhormone-secreting pituitary tumors.


    PRL-Secreting Pituitary Tumors
 Top
 Introduction
 PRL-Secreting Pituitary Tumors
 GH-Secreting Pituitary Tumors
 ACTH-Secreting Pituitary Tumors
 Glycoprotein Hormone (TSH, LH,...
 Summary
 References
 
PRL is the hormone most frequently secreted in excess by pituitary tumors. Although very high PRL levels (>250 ng/mL) are virtually always indicative of a prolactinoma, other causes of hyperprolactinemia need to be considered with more modest elevations of serum PRL levels (1). Physiological elevations of PRL are seen normally during pregnancy and lactation, as well as during sleep and with stress. A variety of drugs, most notably dopamine receptor antagonists such as the neuroleptics and metoclopramide, often cause hyperprolactinemia. Elevated PRL levels also occur in primary hypothyroidism and in renal failure. In the absence of these conditions, however, hyperprolactinemia usually indicates a hypothalamic or pituitary disorder, but not necessarily a prolactinoma. Any sellar or parasellar process that compresses the pituitary stalk and interferes with tonic dopamine inhibition of PRL secretion can lead to hyperprolactinemia. This is an important consideration therapeutically because the treatment of patients with large pituitary masses and moderately elevated PRL levels is quite different than the treatment of patients with large prolactinomas. GH-secreting tumors may also hypersecrete PRL. PRL levels are usually greater than 250 ng/mL in patients with large prolactinomas and are usually less than 100 ng/mL in patients with pituitary stalk compression. PRL levels may be only moderately elevated (<100 ng/mL) in patients with PRL-secreting microadenomas, defined as less than 10 mm in size. Patients without any obvious cause of hyperprolactinemia and with normal high resolution MRI scans are classified as having idiopathic hyperprolactinemia. Some of these patients may actually have very small prolactinomas, but the etiology of the hyperprolactinemia in many of these patients remains unclear.

Patients with PRL-secreting microadenomas usually present with symptoms caused by the high PRL levels. Patients with PRL-secreting macroadenomas, however, often present with additional symptoms related to the presence of a pituitary mass lesion. The treatment of patients with prolactinomas varies with the symptoms of the patient and the size of the tumor (2). Hyperprolactinemia causes hypogonadotropic hypogonadism in women and men, primarily due to the inhibitory effect of high PRL levels on hypothalamic GnRH release. The most common symptoms of hyperprolactinemia are amenorrhea and galactorrhea in women and decreased libido and impotence in men. Treatment is clearly indicated to restore fertility and to reverse symptomatic hypogonadism. The long-term effects of hypogonadism on bone mineral density and the cardiovascular system also need to be considered. A major therapeutic objective is also to reduce the size of the tumor and to limit future growth. Reduction in tumor size is of primary importance in patients with large macroadenomas and symptoms of a mass lesion, including visual symptoms secondary to compression of the optic chiasm, cranial neuropathies, and hypopituitarism. This is not the case with microadenomas. Several studies now support the view that in the majority of patients, untreated microadenomas do not grow into macroadenomas (2, 3). Thus, the presence of a small tumor per se (in the absence of clinical symptoms) is not a definite indication for surgical intervention or for medical intervention. These patients must, however, continue to be carefully monitored. The presence of a macroadenoma, however, even without symptoms, usually warrants treatment to prevent future growth.

Medical treatment with long-acting dopamine agonists is very effective in reducing PRL levels and restoring gonadal function in patients with prolactinomas. In addition, these drugs cause significant tumor shrinkage in ~75% of patients (2, 4). In some patients with large invasive tumors, the shrinkage can be dramatic. Changes in visual fields can be noted within days after initiating therapy. However, if the dopamine agonist is stopped, hyperprolactinemia usually recurs as well as reexpansion of the tumor. Transsphenoidal surgery by an experienced pituitary neurosurgeon is also very effective in curing 70–90% of patients with PRL-secreting microadenomas (4, 5, 6, 7). Initial enthusiasm for transsphenoidal surgery of PRL-secreting microadenomas has, however, been dampened considerably by high recurrence rates (17–50%) as well as by the effectiveness of current medical therapy with long-acting dopamine agonists (6, 7). Surgical cure rates are much lower for macroadenomas. Transsphenoidal surgery, however, continues to be used and is effective if dopamine agonists are not tolerated or if they do not work. Surgery may also be indicated for large tumors with PRL levels in the range seen with pituitary stalk compression, raising the suspicion that the lesion may be something other than a prolactinoma. Very rarely, radiotherapy may be required depending on the response to surgery and dopamine agonist treatment.

The majority of patients with PRL-secreting tumors who require therapy can be safely and effectively treated with dopamine agonists. Bromocriptine and pergolide are the two long-acting dopamine agonists that have been available for use in the United States for many years. Recently, another very long-acting dopamine agonist, cabergoline, has become available in the United States. All three drugs are ergoline derivatives. Pergolide is only approved in the United States for the treatment of Parkinson’s disease, but it has been shown to be an extremely safe and effective treatment for prolactinomas (8, 9). In a large, randomized, controlled multicenter trial, pergolide and bromocriptine were shown to be equally effective in lowering PRL levels and causing tumor shrinkage (9). The advantages to the use of pergolide compared with bromocriptine are that it is much more potent, is longer-acting, and is considerably cheaper. Thus, pergolide can usually be administered in one dose of 0.05–0.1 mg at night. Bromocriptine (2.5 mg) is usually administered two to three times a day. Cabergoline is even longer-acting than pergolide and can be administered at doses of 0.5–1 mg once or twice weekly. Like bromocriptine and pergolide, it is very effective in normalizing PRL levels, restoring gonadal function, and in causing tumor shrinkage. Cabergoline was shown to be better tolerated than bromocriptine in a large double-blind comparison of the two drugs (10). In that study, 3% of patients discontinued cabergoline because of drug intolerance as compared with 12% of patients on bromocriptine. In addition, prolactinomas resistant to other dopamine agonists have been shown to respond to cabergoline (11, 12). In a recent large retrospective study of 452 patients with pathological hyperprolactinemia, most of whom had pituitary tumors, cabergoline was shown to be effective in many patients who were previously bromocriptine intolerant or resistant (13).

Our current practice is to treat most patients who do not wish to become pregnant with cabergoline. Women who wish to become pregnant are still treated initially with bromocriptine because of the extensive safety record with bromocriptine for this purpose. A growing number of women, however, have become pregnant while taking cabergoline and have delivered healthy children, but the numbers are still relatively small. It is anticipated that in the future cabergoline may be recommended as the drug of choice for this purpose as safety data accumulates. Once pregnancy is confirmed, bromocriptine is usually stopped. Women with microadenomas are very unlikely (<5%) to experience problems with clinically significant tumor growth during pregnancy (14). With macroadenomas, however, it is more likely (15–35%) that the tumor will re-expand after bromocriptine withdrawal and may enlarge further, causing compressive symptoms with continued estrogen stimulation during the course of pregnancy. If the patient becomes symptomatic, bromocriptine can be restarted or surgery can be considered. In all patients, when initiating therapy with any of the dopamine agonist drugs, it is important to start with a very small dose to minimize side effects. The most common side effects are nausea, vomiting, and postural hypotension. These side effects usually disappear with continued treatment. With bromocriptine and pergolide it is our practice to begin with a quarter to half tablet once a day with a snack at bedtime. The dose is then gradually increased as tolerated by the patient to 2.5 mg po twice or three times a day for bromocriptine and 0.05–0.1 mg po daily for pergolide. Cabergoline is started at a dose of 0.25 mg po once a week and gradually increased to 0.5 mg once or twice a week. A serum PRL level is checked after 1 and 2 months, and, if necessary, the dose can be increased.

Dopamine agonists should be considered the first line of therapy, even with very large invasive PRL-secreting tumors that are causing compressive symptoms. Medical therapy often causes dramatic tumor shrinkage in these patients, and the vast majority never require surgery. If vision is compromised, visual acuity and visual fields should be carefully monitored. PRL levels should also be measured, and the dose of dopamine agonist should be adjusted accordingly. In general, if the PRL level does not fall, it is unlikely that there will be significant tumor shrinkage. A fall in PRL, however, is not always accompanied by tumor shrinkage. It is, therefore, important to document tumor shrinkage by repeating a MRI in 3–6 months or sooner if vision is impaired and does not improve. In many patients, visual abnormalities may improve rapidly as the tumor shrinks. In some patients with long-standing visual deficits, however, vision may not improve despite tumor shrinkage and decompression of the optic chiasm. In these patients it is unlikely that vision will improve with subsequent surgery. If the optic chiasm is not adequately decompressed, however, surgery may be indicated to restore vision. The majority of patients with large prolactinomas, however, can be adequately treated with dopamine agonists and never require surgery. Once PRL levels fall and there has been significant tumor shrinkage, it is very unlikely that the patient will become resistant to drug therapy. It is not uncommon for the maintenance dopamine agonist dose to be reduced during long-term treatment.

Because the natural history of untreated PRL-secreting microadenomas is quite benign, indications for treatment depend on the patient’s symptoms. Dopamine agonist treatment will correct the hypogonadism and restore fertility in most cases. Alternatively, estrogen replacement therapy can be used to treat symptoms of estrogen deficiency and to prevent the long-term complications of estrogen deficiency. Oral contraceptives can be used to treat younger women who desire contraception. Treatment with estrogen without concurrent treatment with dopamine agonists seems to be safe and not to be associated with significant tumor growth in most cases (15). Patients should be followed regularly, however, and PRL levels should be measured because there is a small risk that estrogen could stimulate tumor growth. If tumor growth is noted while taking estrogen, therapy with a dopamine agonist should be initiated.

In summary, the majority of patients with PRL-secreting microadenomas and macroadenomas can be effectively treated with dopamine agonists. Transsphenoidal surgery is also an effective option for patients who are resistant to or intolerant of these drugs.


    GH-Secreting Pituitary Tumors
 Top
 Introduction
 PRL-Secreting Pituitary Tumors
 GH-Secreting Pituitary Tumors
 ACTH-Secreting Pituitary Tumors
 Glycoprotein Hormone (TSH, LH,...
 Summary
 References
 
GH-secreting pituitary tumors comprise up to 16% of pituitary tumors presenting for surgical therapy (16). These tumors, which are the etiology of acromegaly in almost all cases, present typically after years of symptoms that include acral enlargement, arthropathy, hyperhidrosis, changes in facial features, soft tissue swelling, and symptoms in some cases of pituitary tumor mass effect such as headache, visual changes, or hypopituitarism (17). Once suspected clinically, acromegaly is diagnosed by the finding of an elevation of IGF-I level, failure of adequate GH suppression after oral glucose, and a pituitary mass on MRI. Most cases will be accompanied by persistently elevated GH levels, but a "normal" random GH level does not exclude the diagnosis. Very rarely, ectopic GHRH production can produce acromegaly and should be considered and investigated in patients with biochemical evidence of acromegaly, but absence of a pituitary mass on MRI (17).

The goals of therapy for acromegaly are to both normalize the excess secretion of growth hormone and IGF-I, thereby alleviating the associated symptoms and to surgically remove or debulk large tumors to prevent damage to parasellar structures. Long-term excess GH secretion in acromegaly is associated with a 2-fold greater mortality rate than in the general population and an increased morbidity due to the cardiovascular, pulmonary (airway obstruction and sleep apnea), and malignant complications of this disease (17). Recent evidence suggests that the increased mortality rate associated with persistent active disease can be reduced to normal with therapy that significantly reduces the GH and IGF-I levels (18, 19, 20, 21). Therefore, the goal of therapy should be to achieve adequate biochemical remission as documented by both a normal age-adjusted IGF-I level and by a nadir GH level after oral glucose of less than 1.0 ng/mL by a sensitive GH assay (22).

Transsphenoidal surgery, with success in many patients and very low morbidity, is generally accepted to be the initial mode of therapy for most patients with acromegaly. In recent series from experienced pituitary surgeons, ~90% of microadenomas and half of macroadenomas will achieve biochemical remission based on normalization of IGF-I levels after transsphenoidal surgery alone (21, 23). Nearly two thirds of noninvasive macroadenomas can be cured surgically (23). The recurrence rate after strictly defined biochemical cure is quite low at less than 6% (21, 23). Complications after transsphenoidal surgery occur in less than 7% of cases and include diabetes insipidus, syndrome of inappropriate secretion of antidiuretic hormone, and rarely cerebrospinal fluid leaks or meningitis (23). For those patients who do not achieve biochemical remission after transsphenoidal surgery, currently available therapeutic options include medical therapy with dopamine agonists and/or somatostatin analogs and radiation therapy.

Recent data suggest that radiotherapy may not be as effective in curing acromegaly as previously thought if normalization of IGF-I values is used to define cure. Several recent series have reported that despite a reduction in GH levels in most patients, from only 5–42% of patients will achieve normalization of their IGF-I level after radiotherapy (21, 24, 25). Long-term follow-up comparing various modalities such as gamma knife therapy, proton beam, and focused conventional RT are needed before full assessment of their comparative efficacy can be made. Currently, with the availability of effective medical therapies and the drawbacks of radiotherapy [i.e., the long lag time to therapeutic effect and high incidence (up to 50%) of hypopituitarism with time,] we usually reserve radiotherapy for large or invasive macroadenomas that could not be removed completely surgically or for patients resistant to medical therapy.

Medical therapy has assumed a more prominent role in the treatment of acromegaly as the numbers and efficacy of the therapies available have increased. The two classes of medication currently available in the United States for therapy of acromegaly are dopamine agonists and somatostatin analogs. Although many acromegalics treated with the dopamine agonist bromocriptine will have some amelioration in symptoms, only 20% of patients will achieve a GH less than 5 ng/mL and IGF-I level will normalize in only 10% (26). Bromocriptine is often required in large doses (up to 20 mg/day), and side effects are common. Cabergoline, the new more potent dopamine agonist, approved in the United States for use in PRL-secreting tumors, holds promise for improved efficacy and tolerability over bromocriptine for the treatment of acromegaly. In a recent study, in 64 patients with acromegaly treated with cabergoline at doses ranging from 1.0–1.75 mg/week for up to a 40-month period (27), IGF-I levels normalized in 39% of patients and fell to 300–450 µg/L in another 28% (27). Higher basal IGF-I or GH levels were associated with a reduced efficacy. As expected, tumors that cosecrete PRL are more likely to be responsive to cabergoline therapy. As with other dopamine agonists, the most frequent side effects are nausea, constipation, headache, and dizziness; but in comparative studies with bromocriptine in patients with prolactinomas the side effects have been somewhat less with cabergoline (10). In patients with persistent mild to moderate disease after surgery, and who are minimally symptomatic, a trial of a dopamine agonist such as cabergoline should be considered. Cabergoline is begun at a dose of 0.25 mg/week and can be increased as tolerated up to 3.0 mg/week as needed to normalize IGF-I levels.

The most effective medical therapies for acromegaly currently available are the somatostatin analogs. Two forms of the somatostatin analog octreotide are now available in the United States for treatment of acromegaly, a shorter-acting form and a long-acting depot form. The shorter-acting form of octreotide is given as three sc injections daily of 100–250 µg each. Octreotide also can be infused sc via a pump with a reportedly somewhat greater efficacy (28). Doses over 800 µg/day are generally not felt to increase effectiveness (17). The long-acting depot formulation recently released in the United States, Sandostatin LAR Depot (Novartis Pharmaceuticals, East Hanover, NJ), is administered as a monthly im injection in doses ranging from 10 mg to 30 mg (20 mg/month is the recommended starting dose). For patients new to these medications, an initial 2-week trial of the sc formulation of octreotide should be administered to ensure drug tolerability before the first depot injection.

Overall, clinical trials have shown similar control of GH and IGF-I levels in patients switched from the sc form to the depot form of octreotide. Normalization of IGF-I levels can be expected to occur in 64% of patients treated with sc octreotide for up to 3 yr (29), and up to 67% of patients will achieve normalization of IGF-I levels with the depot formulation (30, 31). Although only a small percentage of patients who responded sub-optimally to the sc octreotide had a more favorable response to the depot formation in clinical trials, the improved compliance expected with the depot form may result in a greater observed efficacy in clinical practice with this formulation. Although GH and IGF-I levels do not normalize in up to one third of patients treated with octreotide, symptomatic relief can be expected in most (29), and headache, in particular, is relieved in up to 95% of patients. Recent evidence of improved cardiac functional status with octreotide treatment supports the belief that effective medical therapy may impact on morbidity and mortality in acromegaly (32). One drawback to the use of these medications alone in some patients with acromegaly is that long-term data on tumor shrinkage with their use is scant. Overall, only ~40% of patients treated with octreotide demonstrate a significant decrease in tumor size (33). Tumor shrinkage data with the depot formulation is available in only a very small number of patients.

Both medications are not infrequently associated with abdominal discomfort, loose stools, nausea and mild malabsorption, which in most patients subside with continued use. Although up to 26% of patients will develop new gallstones during somatostatin analog therapy, they remain asymptomatic in the majority of patients and should be managed as would gallstones in nonacromegalic patients (33). In asymptomatic patients, routine gallbladder sonogram during therapy, therefore, is not felt to be necessary (33). Other less frequent side effects include asymptomatic bradycardia, abnormalities of thyroid function, transient hair loss and vitamin B12 deficiency, and also pain at the site of injection in a small percentage of patients receiving the depot formulation (30, 31). Concurrent somatostatin analog and dopamine agonist therapy may also provide additional benefit (34).

The longer-acting depot form of octreotide, because of ease of administration, likely will be chosen as the adjuvant therapy of choice in most patients who fail surgery for acromegaly. The shorter-acting formulation may be preferred in certain situations, such as for preoperative treatment in patients with signs of upper airway compromise or significant cardiovascular disease to attempt to rapidly lower GH levels and possibly reduce perioperative complications (35). In selected cases in which the chance of surgical cure is low or in whom surgery is contraindicated, the somatostatin analogs have been suggested as first line therapy for acromegaly (33, 36). Long-term follow-up of such patients will be needed to compare their outcome, with respect to tumor size in particular, with similar patients receiving transsphenoidal surgery.

In the near future, analogs of GH that act at the GH receptor as antagonists may also enter the armamentarium of drugs available for the treatment of acromegaly. Early data from clinical trials with a GH antagonist, pegvisomant (B2036-PEG), demonstrate that this medication is well tolerated and in sufficient doses normalizes IGF-I levels in nearly all patients with acromegaly, including those resistant to somatostatin analogs (37).

In summary, transsphenoidal surgery is the primary treatment of choice in the majority of patients with acromegaly. The role of primary medical therapy with somatostatin analogs in patients unlikely to be cured surgically is yet to be determined. For those patients with persistent elevation of GH or IGF-I levels after surgery, medical therapy should be undertaken. For those patients with mild disease, a trial of the oral dopamine agonist cabergoline can be initiated. If this fails or in patients with moderate to severe disease, therapy with a somatostatin analog should be begun. In the near future, the GH antagonist pegvisomant may also be another option for medical therapy. Radiotherapy, in conjunction with appropriate medical therapy, should be considered for patients with significant residual tumor after surgery or in whom medical therapy is unsuccessful.


    ACTH-Secreting Pituitary Tumors
 Top
 Introduction
 PRL-Secreting Pituitary Tumors
 GH-Secreting Pituitary Tumors
 ACTH-Secreting Pituitary Tumors
 Glycoprotein Hormone (TSH, LH,...
 Summary
 References
 
Patients with ACTH-secreting pituitary tumors, termed Cushing’s disease, usually present with clinical features resulting from prolonged overproduction of glucocorticoids. These include central obesity, cutaneous atrophy, easy bruisability, muscle wasting, osteoporosis, hypertension, diabetes mellitus, and psychiatric symptoms. In addition, excess production of androgens in women is associated with hirsutism, acne, and amenorrhea. Cushing’s disease occurs five times more frequently in women than in men, with the peak incidence occurring between 20–50 yr of age. Cushing’s disease must be differentiated from other causes of Cushing’s syndrome, which include adrenal tumors and ectopic secretion of ACTH by a nonpituitary tumor, as well as from pseudo-Cushing’s states (38, 39, 40). The diagnosis of Cushing’s disease is usually made on the basis of the clinical and laboratory data rather than by the radiographic localization of the pituitary tumor. Because of the small size of these tumors, 40–50% are not localized preoperatively by MRI with gadolinium-diethyl-enetriaminepentaacetic acid enhancement. It should be emphasized that because a fair number of asymptomatic nonsecreting microadenomas exist in the general population, these radiographic techniques are only helpful once a clinical and biochemical diagnosis of Cushing’s disease has been made. Urine-free cortisol measurements, low-dose dexamethasone testing, and late night plasma and salivary cortisol levels are helpful in establishing a diagnosis of Cushing’s syndrome (39, 40). Plasma ACTH levels, high-dose dexamethasone testing, and CRH stimulation testing can then be used to establish the cause of Cushing’s syndrome. The majority of patients with Cushing’s disease have plasma ACTH levels in the high normal range, but inappropriately elevated for the level of plasma cortisol. In contrast, patients with adrenal tumors have suppressed ACTH levels (<10 pg/mL) that do not respond to stimulation with CRH. Once a diagnosis of ACTH-dependent Cushing’s syndrome has been made it is important to distinguish Cushing’s disease from ectopic ACTH secretion. Patients with ectopic ACTH production may have extremely elevated plasma ACTH levels, but often there is considerable overlap with the ACTH levels measured in Cushing’s disease. CRH testing can be helpful in that most patients with Cushing’s disease respond to CRH administration with a rise in plasma ACTH and cortisol levels, whereas patients with ectopic ACTH secretion usually do not respond to CRH. The most reliable way, however, to differentiate pituitary from ectopic ACTH secretion is by inferior petrosal sinus sampling, which can be used to measure ACTH levels in the venous blood draining the pituitary (41, 42). We routinely perform inferior petrosal sinus sampling in conjunction with CRH stimulation when the MRI is negative and/or standard biochemical studies are inconclusive. This technique also has been reported in some cases to be useful for preoperative lateralization of ACTH-secreting pituitary adenomas and, thus, for guiding the surgeon in performing a partial hypophysectomy if no adenoma is seen.

Transsphenoidal surgery is currently the treatment of choice for patients with Cushing’s disease. Microadenomas are usually found in most patients at the time of surgery and can be selectively removed with eventual resumption of normal hypothalamic-pituitary-adrenal function. If no adenoma is found, a partial hypophysectomy guided by the petrosal sinus ACTH levels may be elected. At times, in an older patient in whom fertility is not an issue, a near total hypophysectomy may be elected because of the morbidity and mortality associated with this disease. In several large series of patients undergoing transsphenoidal surgery for Cushing’s disease, cure rates of 80–90% have been reported for tumors confined to the sella (43, 44). Most patients have been cured by selective adenomectomy, but a small number have been cured by total hypophysectomy if no tumor was identified at surgery. In a large series reported by Mampalam et al. (43), there was an overall remission rate of 76% in 216 patients after transsphenoidal surgery. In that series, 9 of 164 patients had recurrence of disease. The disease recurred at an average of 3.8 yr after surgery.

Postoperatively, if the patient has been cured, there will be transient hypoadrenalism until the normal hypothalamic-pituitary-adrenal axis recovers. If hypocortisolism is not demonstrable after surgery, it is unlikely that the patient is cured. The best long-term prognosis for cure is seen in patients with plasma cortisol levels of less than 1µg/dL who require prolonged glucocorticoid replacement (45). After successful transsphenoidal surgery, patients will typically require daily glucocorticoid replacement for 3–12 months (46). Cured patients often complain of steroid withdrawal symptoms, including myalgias and arthralgias, despite physiological glucocorticoid replacement. Physicians should be aware that previous psychopathology may persist into the postoperative period and new psychopathology may emerge. In one series, 66% of patients with Cushing’s syndrome had significant psychopathology, with a predominant diagnosis of atypical depression; after the correction of hypercortisolism, overall psychopathology decreased but the frequency of suicidal ideation and panic increased (47).

If the patient is not cured by surgery, options include a repeat transsphenoidal procedure, radiotherapy or, rarely, bilateral adrenalectomy. Depending on the extent of the first surgical procedure, a second more extensive hypophysectomy may be performed. Pituitary radiotherapy should then be considered as the next most appropriate treatment for patients not cured by transsphenoidal surgery. Although conventional pituitary irradiation has been successful in curing childhood Cushing’s disease, much lower cure rates have been reported in adults when used as primary therapy. Pituitary irradiation has, however, recently been shown to be an effective treatment in patients who have had transsphenoidal surgery but were not cured (48). In that series, 83% of patients had remissions after radiotherapy. The remissions began 6–60 months after radiation therapy, but in most cases remission occurred within 2 yr of treatment. The only side effect noted in this series was the development of variable degrees of hypopituitarism in 15 of 30 patients. Newer forms of stereotactic radiotherapy (a computer-assisted linear accelerator or cobalt-60, the gamma knife) may prove to be very effective, but long-term follow-up with these techniques for treating pituitary tumors is limited. Medical therapy is usually required to lower cortisol levels while waiting for the radiotherapy to take effect.

Medical therapy for Cushing’s disease is aimed primarily at the adrenal and consists of several inhibitors of adrenal steroidogenesis (49). Ketoconazole, an imidazole derivative that inhibits cortisol synthesis at multiple steps, has been shown to rapidly suppress cortisol levels in patients with Cushing’s disease (49, 50, 51). Ketoconazole inhibits cholesterol side-chain cleavage, 11ß-hydroxylase, and 17{alpha}-hydroxylase; inhibition of the 17, 20 lyase also affects androgen synthesis. The initial oral dose is usually 200 mg administered every 12 h, but doses of up to 1200 mg/day may be required. The major side effects are alterations in hepatic function and gastrointestinal symptoms. Liver function tests should be monitored during therapy. At high doses ketoconazole may impair testicular function. If ketoconazole is not effective or well tolerated, other adrenal enzyme inhibitors including metyrapone, aminoglutethimide, and o,p’DDD (mitotane) can be used. Lower doses of several drugs can be used in combination to minimize side effects. Etomidate, an imidazole derivative related to ketoconazole, has been used iv to lower cortisol levels in patients who cannot take oral medications (52). Mifepristone, a glucocorticoid receptor antagonist, has been used in a small number of patients, but it is difficult to monitor its effectiveness other than by the clinical response because cortisol levels will not fall and may actually increase. Medical therapy is not considered primary therapy for established Cushing’s disease, but is often used in conjunction with radiotherapy or preoperatively in a very debilitated patient to improve the patient’s clinical condition before surgery.

Bilateral adrenalectomy, which was the standard treatment in the past, is occasionally still used in patients who fail surgery and radiotherapy, but life-long treatment with glucocorticoid and mineralocorticoid will be required. In addition, patients are at risk for Nelson’s syndrome manifested by a progressive increase in skin pigmentation due to an increase in ACTH levels and pituitary tumor growth. The risk is reduced, however, by prior pituitary radiotherapy.

In summary, the majority of patients with Cushing’s disease can be cured by transsphenoidal surgery with selective adenomectomy with subsequent recovery of normal hypothalamic-pituitary-adrenal function. Patients will require glucocorticoid replacement and careful monitoring during the period of transient adrenal insufficiency that occurs before the hypothalamic-pituitary-adrenal axis recovers. If surgery fails, a more extensive surgical procedure and/or radiotherapy is indicated. Medical therapy with ketoconazole and other adrenal enzyme inhibitors can be used as adjunctive therapy to lower cortisol levels.


    Glycoprotein Hormone (TSH, LH, FSH)-Secreting and Nonfunctioning Pituitary Tumors
 Top
 Introduction
 PRL-Secreting Pituitary Tumors
 GH-Secreting Pituitary Tumors
 ACTH-Secreting Pituitary Tumors
 Glycoprotein Hormone (TSH, LH,...
 Summary
 References
 
TSH-secreting tumors are very rare, comprising less than 1% of pituitary adenomas. These tumors may present with symptoms of hyperthyroidism, a nonsuppressed TSH level, and signs of a pituitary mass (53, 54). The major differential diagnosis to consider when evaluating patients with thyroid function tests suggestive of a TSH-secreting pituitary tumor is the syndrome of thyroid hormone resistance. In most, but not all cases, the classical criteria (i.e., lack of TSH response to TRH stimulation, elevation of serum {alpha}-subunit, and high {alpha}-subunit/TSH ratio along with a pituitary mass on MRI are diagnostic of these tumors) (54). TSH-secreting pituitary tumors should be treated initially surgically (53, 54). Only one third of these tumors will be cured by surgery alone because most are macroadenomas, which tend to be locally invasive at presentation. Radiotherapy is recommended as routine adjunctive therapy when surgery has not been curative (54). Even after both surgery and radiotherapy only approximately two thirds of tumors will be under control biochemically (53). There is very little success with dopamine agonists for treatment of these tumors. With octreotide therapy, however, biochemical disease can be controlled in most patients; TSH levels have been reported to normalize in 79% of patients, and tumor shrinkage can occur in 52% of patients treated with octreotide (53, 54).

FSH- and LH-secreting tumors that present with clinical symptoms of hormone hypersecretion are very rare (55). Thus, the majority of gonadotroph-producing tumors are clinically hormonally silent. Pituitary tumors that present without clinical and biochemical evidence of pituitary hormone hypersecretion are termed clinically nonfunctioning adenomas and are in the majority of cases of gonadotroph origin. Although clinically silent, in vitro evidence demonstrates that most of these tumors do synthesize glycoprotein hormones or their subunits, including FSH, free {alpha} or ß subunits, and, less commonly, LH (55). Serum gonadotropin concentrations are usually normal, but elevations of serum-free ß-FSH or {alpha}-subunit may be detected in some patients (56). Some have found elevation of serum {alpha}-subunit to be a useful marker (57), as well as an increase in gonadotropin or subunit levels after TRH stimulation (58).

Nonfunctioning tumors, usually macroadenomas, typically come to medical attention because of visual disturbance, symptoms of hypopituitarism, or headache (59, 60). These tumors need to be distinguished from other sellar/parasellar masses that can mimic a pituitary adenoma (16). The major considerations in the differential diagnosis of pituitary adenomas are cysts, craniopharyngiomas, meningiomas, metastatic tumors to the pituitary, granulomatous and infiltrative processes, and lymphocytic hypophysitis. Diabetes insipidus is very uncommon at presentation in pituitary adenomas and should prompt consideration of a nonpituitary lesion. Although certain clinical and radiographic features may help in the diagnosis of nonpituitary lesions, some will not be diagnosed until the time of surgery (16). Nonfunctioning tumors may be accompanied by modest hyperprolactinemia, usually less than 100 ng/mL, due to compression of the pituitary stalk.

The initial management of nonfunctioning pituitary macroadenomas is transsphenoidal surgery with the goals being removal of tumor mass and decompression of parasellar structures. Transsphenoidal surgery has a low morbidity, results in improvement in visual loss in many cases (59), and may lead to recovery of pituitary function. Because of the large size and parasellar extension found in many of these tumors, they often cannot be completely resected. Even after apparent resection, recurrence rates after surgery may be as high as 16% (59). Postoperative radiotherapy is an option for treatment of residual tumor. However, clear prospective data on the benefits and outcome of radiotherapy for nonsecretory tumors are not available. The decision about whether to proceed to radiotherapy after surgery depends on the size of the residual tumor, the patient’s age, and the desire to retain pituitary function. Postoperative radiotherapy is clearly not necessary in every case. One may elect to follow with serial scans every 6 months, for the 1st yr, and then yearly thereafter to determine the rate of growth of the tumor and reconsider radiotherapy if there is tumor growth.

Attempts at medical therapy of nonfunctioning pituitary tumors with dopamine agonists and analogs of somatostatin have met with little success, in part because only a minority of these tumors bear dopamine or somatostatin receptors. In a few reported cases, bromocriptine therapy has decreased tumor size, but tumor shrinkage does not occur in most patients. Similarly, use of octreotide has been successful in improving vision in some patients with these tumors (56), but the response is very heterogeneous and its use should be reserved as with bromocriptine for patients unresponsive to other therapies.

Nonfunctioning pituitary tumors are not infrequently discovered on imaging studies done for unrelated reasons in patients without symptoms referable to the pituitary tumor. The management of these asymptomatic tumors or incidentalomas is somewhat controversial. Nonfunctioning microadenomas (<10 mm) need not be treated surgically and can be followed with serial MRI. Most agree, however, that surgery should be considered for asymptomatic macroadenomas. Depending on the size of the tumor and other medical conditions, however, one may elect to monitor the tumor on serial pituitary imaging studies and follow clinical signs and pituitary function. Little prospective data, however, is actually available on the natural history of asymptomatic pituitary tumors.


    Summary
 Top
 Introduction
 PRL-Secreting Pituitary Tumors
 GH-Secreting Pituitary Tumors
 ACTH-Secreting Pituitary Tumors
 Glycoprotein Hormone (TSH, LH,...
 Summary
 References
 
We are fortunate to have multiple safe and effective therapeutic options available for the treatment of pituitary tumors. These options include medical therapy, transsphenoidal surgery and radiotherapy. The treatment of choice depends on the type of pituitary tumor. The majority of PRL-secreting tumors can be effectively treated with dopamine agonists. Transsphenoidal surgery is also an effective option for patients who are resistant to or intolerant of these drugs. Transsphenoidal surgery remains the treatment of choice for the majority of patients with GH, ACTH, and TSH-secreting tumors and for large nonsecreting tumors. Medical therapy with somatostatin analogs and/or dopamine agonists should be undertaken in patients with persistent elevations of GH and IGF-I levels; radiotherapy should be considered for patients with significant residual tumor in whom medical therapy is unsuccessful. Radiotherapy is also indicated for ACTH-secreting tumors not cured by surgery; medical therapy with ketoconazole and other adrenal enzyme inhibitors can be used as adjunctive therapy to lower cortisol levels. Postoperative radiotherapy for nonsecreting tumors is also an option if there is considerable residual tumor or evidence of tumor growth on follow-up MRI. Evaluation and treatment of hypopituitarism is an important part of the management of all patients with pituitary tumors. Patients also should be monitored for the development of new deficits, particularly after radiotherapy. The development of new medical therapies, such as GH antagonists, as well as refinements of surgical, radiotherapy, and imaging techniques should continue to improve our management of pituitary tumors.

Received July 28, 1999.

Revised September 17, 1999.

Accepted September 17, 1999.


    References
 Top
 Introduction
 PRL-Secreting Pituitary Tumors
 GH-Secreting Pituitary Tumors
 ACTH-Secreting Pituitary Tumors
 Glycoprotein Hormone (TSH, LH,...
 Summary
 References
 

  1. Frantz AG. 1988 Hyperprolactinemia. In: Collu R, Brown GM, Van Loon GR, eds. Clinical Neuroendocrinology. Boston: Blackwell Scientific Publications; 311–332.
  2. Molitch ME. 1999 Medical treatment of prolactinomas. Endocrinol Metab Clin North Am. 28:143–169.[Medline]
  3. Schlechte J, Donlan K, Sherman B, Chapler F, Luciano A. 1989 The natural history of untreated hyperprolactinemia: a prospective analysis. J Clin Endocrinol Metab. 68:412–418.[Abstract]
  4. Molitch ME, Thorner MO, Wilson C. 1997 Therapeutic controversy: management of prolactinomas. J Clin Endocrinol Metab. 82:996–1000.[Free Full Text]
  5. Feigenbaum SL, Downey DE, Wilson CB, Jaffe RB. 1996 Transsphenoidal pituitary resection for preoperative diagnosis of prolactin-secreting pituitary adenoma in women: long-term follow-up. J Clin Endocrinol Metab. 81:1711–1719.[Abstract]
  6. Serri O, Rasio E, Beauregard H, Hardy J, Somma M. 1983 Recurrence of hyperprolactinemia after selective transsphenoidal adenomectomy in women with prolactinoma. N Engl J Med. 309:280–283.[Abstract]
  7. Rodman EF, Molitch ME, Post KD, Biller BJ, Reichlin S. 1984 Long-term follow-up of transsphenoidal selective adenomectomy for prolactinoma. JAMA. 252:921–924.[Abstract]
  8. Kleinberg DL, Boyd AE, Wardlaw S, et al. 1983 Pergolide for the treatment of pituitary tumors secreting prolactin or growth hormone. N Engl J Med. 309:704–709.[Abstract]
  9. Lamberts SWJ, Quik RFP. 1991 A comparison of the efficacy and safety of pergolide and bromocriptine in the treatment of hyperprolactinemia. J Clin Endocrinol Metab. 72:635–641.[Abstract]
  10. Webster J, Piscitelli G, Polli A, Ferrari CI, Ismail I, Scanlon MF. 1994 A comparison of cabergoline and bromocriptine in the treatment of hyperprolactinemic amenorrhea. N Engl J Med. 331:904–909.[Abstract/Free Full Text]
  11. Biller BMK, Molitch ME, Vance ML, et al. 1996 Treatment of prolactin-secreting macroadenomas with the once-weekly dopamine agonist cabergoline. J Clin Endocrinol Metab. 81:2338–2343.[Abstract]
  12. Colao A, Di Sarno A, Sarnacchiaro F, et al. 1997 Prolactinomas resistant to standard dopamine agonists respond to chronic cabergoline treatment. J Clin Endocrinol Metab. 82:876–883.[Abstract/Free Full Text]
  13. Verhelst J, Abs R, Maiter D, et al. 1999 Cabergoline in the treatment of hyperprolactinemia: a study in 455 patients. J Clin Endocrinol Metab. 84:2518–2522.[Abstract/Free Full Text]
  14. Molitch ME. 1985 Current concepts: pregnancy and the hyperprolactinemic woman. N Engl J Med. 312:1364–1370.[Medline]
  15. Corenblum B, Donovan L. 1993 The safety of physiological estrogen plus progestin replacement therapy and with oral contraceptive therapy in women with pathological hyperprolactinemia. Fertil Steril. 59:671–673.[Medline]
  16. Freda PU, Post KD, Wardlaw SL. 1996 Unusual causes of sellar/parasellar masses in a large transsphenoidal surgical series. J Clin Endocrinol Metab. 81:3455–3459.[Medline]
  17. Melmed S, Ho K, Klibanski A, Reichlin S, Thorner M. 1995 Clinical review 75: recent advances in pathogenesis, diagnosis and mangement of acromegaly. J Clin Endocrinol Metab. 80:3395–3402.[Medline]
  18. Bates AS, Van’t Hoff W, Jones JM, Clayton RN. 1995 Does treatment of acromegaly affect life expectancy? Metab Clin Exp. 44:1–5.
  19. Rajasoorya C, Holdaway IM, Wrightson P, Scott DJ, Ibbertson HK. 1994 Determinants of clinical outcome and survival in acromegaly. Clin Endocrinol. 41:95–102.[Medline]
  20. Orme SM, McNanny RJQ, Cartwright RA, Belchetz PE. 1998 Mortality and cancer incidence in acromegaly: a retrospective cohort study. J Clin Endocrinol Metab. 83:2730–2734.[Abstract/Free Full Text]
  21. Swearingen B, Barker FG, Katznelson L, et al. 1998 Long-term mortality after transsphenoidal surgery and adjunctive therapy for acromegaly. J Clin Endocrinol Metab. 83:3419–3426.[Abstract/Free Full Text]
  22. Freda PU, Post KD, Powell JS, Wardlaw SL. 1998 Evaluation of disease status with sensitive measures of GH secretion in 60 postoperative patients with acromegaly. J Clin Endocrinol Metab. 83:3808–3816.[Abstract/Free Full Text]
  23. Freda PU, Wardlaw SL, Post KD. 1998 Long-term endocrinological follow-up evaluation in 115 patients who underwent transsphenoidal surgery for acromegaly. J Neurosurg. 89:353–358.[CrossRef][Medline]
  24. Barkan AL, Halasz I, Dornfield KJ, et al. 1997 Pituitary irradiation is ineffective in normalizing plasma insulin-like growth factor I in patients with acromegaly. J Clin Endocrinol Metab. 82:3187–3191.[Abstract/Free Full Text]
  25. Powell JS, Wardlaw SL, Post KD, Freda PU. 1999 Outcome of radiotherapy for acromegaly using normalization of IGF-I level to define cure. Proc 81st Annual Meeting of The Endocrine Society, San Diego, CA, 1999, P3–652 (Abstract).
  26. Jaffe CA, Barkan AL. 1992 Treatment of acromegaly with dopamine agonists. Endocrinol Metab Clin North Am. 21:713–735.[Medline]
  27. Abs R, Verhelst J, Maiter D, et al. 1998 Cabergoline in the treatment of acromegaly: a study in 64 patients. J Clin Endocrinol Metab. 83:374–378.[Abstract/Free Full Text]
  28. Tauber JP, Babin TH, Tauber MT, et al. 1997 Long-term effects of continuous subcutaneous infusion of the somatostatin analog octreotide in the treatment of acromegaly. J Clin Endocrinol Metab. 68:917–924.[Abstract]
  29. Newman CB, Melmed S, Snyder PJ, et al. 1995 Safety and efficacy of long-term octreotide therapy of acromegaly; results of a multicenter trial in 103 patients—a clinical research center study. J Clin Endocrinol Metab. 80:2768–2775.[Abstract]
  30. Flogstad AK, Halse J, Bakke S, et al. 1997 Sandostatin LAR in acromegalic patients: long-term treatment. J Clin Endocrinol Metab. 81:23–28.
  31. Lancranjan I, Atkinson AB, The Sandostatin LAR Group. 1999 Results of a European multicenter study with sandostatin LAR in acromegalic patients. Pituitary. 1:105–114.[CrossRef][Medline]
  32. Colao A, Cuocolo A, Marzullo P, et. al. 1999 Effects of a 1-year treatment with octreotide on cardiac performance in patients with acromegaly. J Clin Endocrinol Metab. 84:17–23.[Abstract/Free Full Text]
  33. Newman CB. 1999 Medical therapy for acromegaly. Endocrinol Metab Clin North Am. 28:171–190.[Medline]
  34. Flogstad AK, Halse J, Grass P, et al. 1994 A comparison of octreotide, bromocriptine, or a combination of both drugs in acromegaly. J Clin Endocrinol Metab. 79:461–465.[Abstract]
  35. Colao A, Ferone D, Cappabianca P, et al. 1997 Effect of octreotide pretreatment on surgical outcome in acromegaly. J Clin Endocrinol Metab. 82:3308–3314.[Abstract/Free Full Text]
  36. Melmed, S, Jackson I, Kleinberg D, Klibanski A. 1998 Current treatment guidelines for acromegaly. J Clin Endocrinol Metab. 83:2646–2652.[Abstract/Free Full Text]
  37. Trainer PJ, Besser GM, Klibanski A, Freda PU, Melmed S, The Sensus Acromegaly Study Group. 1999 A phase III study of B2036-PEG, a growth hormone antagonist, in the treatment of acromegaly. Proc 81st Annual Meeting of The Endocrine Society, San Diego, CA, 1999 (Abstract P1-46).
  38. Orth DN. 1995 Cushing’s syndrome. N Engl J Med. 332:791–803.[Free Full Text]
  39. Newell-Price J, Trainer P, Besser M, Grossman A. 1998 The diagnosis and differential diagnosis of Cushing’s syndrome and pseudo-Cushing’s states. Endocr Rev. 19:647–672.[Abstract/Free Full Text]
  40. Findling JW, Raff H. 1999 Newer diagnostic techniques and problems in Cushing’s disease. Endocrinol Metab Clin North Am. 28:191–210.[Medline]
  41. Oldfield EH, Doppman JL, Nieman LK, et al. 1991 Petrosal sinus sampling with and without corticotropin-releasing hormone for the differential diagnosis of Cushing’s syndrome. N Engl J Med. 325:897–905.[Abstract]
  42. Kaltsas GA, Giannulis MG, Newell-Price JDC, et al. 1999 A critical analysis of the value of simultaneous inferior petrosal sinus sampling in cushing’s disease and the occult ectopic adrenocorticotropin syndrome. J Clin Endocrinol Metab. 84:487–492.[Abstract/Free Full Text]
  43. Mampalam TJ, Tyrrell B, Wilson CB. 1988 Transsphenoidal microsurgery for Cushing disease: a report of 216 cases. Ann Intern Med. 109:487–493.[Medline]
  44. Swearingen B, Biller BMK, Barker FG II, et al. 1999 Long-term mortality after transsphenoidal surgery for Cushing disease. Ann Intern Med. 130:821–824.[Abstract/Free Full Text]
  45. Bochicchio D, Losa M, Buchfelder M. 1995 Factors influencing the immediate and late outcome of cushing’s disease treated by transsphenoidal surgery: a retrospective study by the European Cushing’s disease survey group. J Clin Endocrinol Metab. 80:3114–3120.[Abstract]
  46. Gomez MT, Magiakou MA, Mastorakos G, Chrousos GP. 1993 The pituitary corticotroph is not the rate-limiting step in the postoperative recovery of the hypothalamic-pituitary-adrenal axis in patients with Cushing’s syndrome. J Clin Endocrinol Metab. 77:173–177.[Abstract]
  47. Dorn LD, Burgess ES, Friedman TC, Dubbert B, Gold PW, Chrousos GP. 1997 The longitudinal course of psychopathology in Cushing’s syndrome after correction of hypercortisolism. J Clin Endocrinol Metab. 82:912–919.[Abstract/Free Full Text]
  48. Estrada J, Boronat M, Mielgo M, et al. 1997 The long-term outcome of pituitary irradiation after unsuccessful transsphenoidal surgery in Cushing’s disease. N Engl J Med. 336:172–177.[Abstract/Free Full Text]
  49. Sonino N, Boscaro M. 1999 Medical therapy for Cushing’s disease. Endocrinol Metab Clin North Am. 28:211–222.[Medline]
  50. Sonino N. 1987 The use of ketoconazole as an inhibitor of steroid production. N Engl J Med. 317:812–818.[Medline]
  51. Loli P, Berselli ME, Tagliaferri M. 1986 Use of ketoconazole in the treatment of Cushing’s syndrome. J Clin Endocrinol Metab. 63:1365–1371.[Abstract]
  52. Drake WM, Perry LA, Hinds CJ, Lowe DG, Reznek RH, Besser GM. 1998 Emergency and prolonged use of intravenous etomidate to control hypercortisolemia in a patient with Cushing’s syndrome and peritonitis. J Clin Endocrinol Metab. 83:3542–3544.[Abstract/Free Full Text]
  53. Beck-Peccoz P, Bruckner-Davis F, Persani L, Smallridge RC, Weintraub BD. 1996 Thryrotropin-secreting pituitary tumors. Endocrine Rev. 17:610–638.[Medline]
  54. Brucker-Davis F, Oldfield EH, Skarulis MC, Doppman JL, Weintraub BD. 1999 Thryotropin-secreting piuitary tumors: diagnostic criteria, thyroid hormone sensitivity, and treatment outcome in 25 patients followed at the National Institutes of Health. J Clin Endocrinol Metab. 84:476–486.[Abstract/Free Full Text]
  55. Katznelson L, Alexander JM, Klibanski A. 1993 Clinical review 45: clinically nonfunctioning pituitary adenomas. J Clin Endocrinol Metab. 76:1089–1094.[Medline]
  56. Shomali ME, Katznelson L. 1999 Medical therapy for gonadotroph and thyrotroph tumors. Endocrinol Metab Clin North Am. 28:223–240.[Medline]
  57. Oppenheim DS, Klibanski A. 1989 Medical therapy of glycoprotein hormone-secreting pituitary tumors. Endocrinol Metab Clin North Am. 18:339–358.[Medline]
  58. Lamberts SWJ, de Herder WW, van der Lely AJ, Hofland LJ. 1995 Imaging and medical management of clinically nonfunctioning pituitary tumors. Endocrinologist. 5:448–451.
  59. Ebersold MJ, Quast LM, Laws ER, Scheithauer B, Randall RV. 1986 Long-term results in transsphenoidal removal of nonfunctioning pituitary adenomas. J Neurosurg. 64:713–719.[Medline]
  60. Snyder PJ. 1993 Clinically nonfunctioning pituitary adenomas. Endocrinol Metab Clin North Am. 22:163–175.[Medline]