TSH-secreting pituitary adenomas

INTRODUCTION — Thyroid-stimulating hormone (TSH)-secreting pituitary adenomas are a rare cause of hyperthyroidism [1]. They account for 0.5 to 3 percent of all functioning pituitary tumors [2] and much less than 1 percent of all cases of hyperthyroidism. The incidence in Sweden is calculated to be 2.8 per 1 million, in which 0.85 per million had disease [1]. Nevertheless, the diagnosis should be considered in all hyperthyroid patients, especially those with a diffuse goiter and no extrathyroidal manifestations of Graves' disease.

This topic will review the clinical presentation, diagnosis, and treatment of TSH-secreting pituitary tumors. Other causes of hyperthyroidism are reviewed separately. (See "Disorders that cause hyperthyroidism" and "Diagnosis of hyperthyroidism".)

PATHOPHYSIOLOGY — TSH-secreting adenomas secrete biologically active TSH in a more or less autonomous fashion. Thus, TSH secretion usually does not increase much in response to thyrotropin-releasing hormone (TRH) and does not decrease much in response to exogenous thyroid hormone administration. The biological activity of the TSH that is secreted varies considerably; as a result, serum immunoreactive TSH concentrations range from normal (albeit inappropriately high in the presence of hyperthyroidism) to markedly elevated (>500 mU/L) [3].

Most TSH-secreting adenomas secrete only TSH. However, approximately 20 to 30 percent of the adenomas secrete one or more other pituitary hormones, predominantly growth hormone or prolactin [2,4]. There have been no reported instances of cosecretion of corticotropin (ACTH) and TSH.

Adenomas secreting TSH and growth hormone are equally common in males and females, whereas cosecretion of TSH and prolactin is approximately five times more common in females than in males. Hyperprolactinemia is not always due to tumor secretion of prolactin; in some patients, it is caused by compression of the pituitary stalk and interruption of tonic hypothalamic inhibition of prolactin secretion. (See "Causes of hyperprolactinemia", section on 'Decreased dopaminergic inhibition of prolactin secretion'.)

The molecular basis of TSH-secreting adenomas is not known. While somatic mutations or abnormal oncogene expression could contribute to these tumors as in the case of other pituitary tumors, no such mutations have been uniformly described. The absence of negative feedback regulation of thyroid hormone on TSH suggests impairment of the thyroid hormone receptor beta 1 (THRB1) or beta 2 (THRB2) gene. In the TSH-secreting adenomas that are plurihormonal, an isolated thyroid hormone receptor beta (THRB) abnormality would not be the etiology. Furthermore, the relationship between tumor growth and hormone secretion is unclear. However, molecular analysis of TSH-secreting adenomas has shown:

●Overexpression of the pituitary-specific transcription factor-1 (Pit-1) in 14 patients [5-7].

●Somatic mutations of the thyroid hormone receptor beta (THRB) gene in two of six TSH-secreting adenomas [8,9]. In the latter, the mutant thyroid hormone receptor beta protein (TR-beta) did not bind thyroid hormone and interfered with the function of the normal TR-beta, thus providing an explanation for the nonsuppressible nature of at least some of the adenomas.

●Decreased expression of TR-beta and TR-alpha in two TSH-secreting adenomas [10] and decreased expression of the corepressor NCOR2 in a single case [11] have been reported.

Activating mutations in the TRH-receptor gene have not been identified [12].

CLINICAL PRESENTATION

Symptoms and signs — Most patients have the typical symptoms and signs of hyperthyroidism (eg, palpitations, tremor, heat intolerance), but a few patients have mild or even no hyperthyroid symptoms [4,13,14] (see "Overview of the clinical manifestations of hyperthyroidism in adults"). In addition, patients may have symptoms related to the expanding tumor mass with compression of the normal pituitary gland or the optic chiasm or from cosecretion of growth hormone or prolactin.

In two separate reviews of patients with TSH-secreting tumors, clinical features other than hyperthyroidism included [3,4,15]:

●A diffuse goiter – 56 to 93 percent

●Visual field defects – 25 to 35 percent

●Menstrual disturbances – 33 percent

●Galactorrhea (women), with or without cosecretion of prolactin – 28 percent

●Headache – 21 percent

In cases of mixed TSH/growth hormone-secreting tumors, symptoms of acromegaly may also occur, including macroglossia, deepening of the voice, carpal tunnel syndrome, and hyperhidrosis. (See "Causes and clinical manifestations of acromegaly".)

In a case series from a single institution in Italy of 62 patients with TSH-secreting pituitary adenoma, three patients had coexisting differentiated thyroid cancer [16], and in an earlier literature review, nine patients with TSH-secreting pituitary adenomas were found to have differentiated thyroid cancer [17]. This finding may possibly be due to long-term TSH stimulation. In most cases, differentiated thyroid cancer was diagnosed and thyroidectomy performed before the diagnosis of a TSH-secreting pituitary adenoma was appreciated.

The characteristic signs of Graves' ophthalmopathy (proptosis and periorbital edema) are absent, unless there is coexisting Graves' disease [18,19].

In two series, the mean age at presentation was 41 to 45 years and equal incidence in males and females [3,20]. In contrast, Graves' disease and most other causes of hyperthyroidism are much more common in women. Additional reports of Graves' disease concurrent with a TSH-secreting adenoma [21,22] may reflect presence of a common disease with a rare disease. The time from onset of symptoms to diagnosis of the pituitary tumor has ranged from 1 to 27 years. Most of the longstanding cases erroneously received ablative thyroid treatment at least once and often several times before their adenomas were correctly diagnosed. In such cases, patients whose thyroid glands were ablated may present as having high serum TSH levels that do not respond to escalating doses of T4 (levothyroxine) replacement.

Thyroid function tests — The characteristic biochemical abnormalities in patients with hyperthyroidism caused by a TSH-secreting adenoma are normal or high serum TSH concentrations and high serum total and free thyroxine (T4) and triiodothyronine (T3) concentrations.

Among 255 reviewed patients [3], the ranges of hormone values were:

●Serum TSH concentrations – <1.0 to 568 mU/L

●Serum total T4 concentrations – 11.6 to 53 micrograms/dL (150 to 678 nmol/L)

●Serum free T4 concentrations – 1.6 to 7.7 ng/dL (20 to 100 pmol/L)

●Serum total T3 concentrations – 195 to 1300 ng/dL (3 to 20 nmol/L)

●Serum free T3 concentrations – 5 to 26 pg/mL (8.0 to 40.2 pmol/L)

Approximately 30 percent of patients who had not had any thyroid ablation therapy in the past had serum TSH values within the normal range; however, "normal" values are inappropriately high in the presence of high serum T4 and T3 concentrations. Frequent measurements of serum TSH in four patients demonstrated loss of the normal nocturnal surge in TSH secretion, a finding that is compatible with autonomous TSH secretion [23]. Serum TSH concentrations do not increase in response to thyrotropin-releasing hormone (TRH) in the majority of patients [3,24].

Alpha subunit — Approximately 50 to 85 percent of patients with TSH-secreting pituitary adenomas (particularly macroadenomas) have a high serum concentration of the alpha subunit of glycoprotein hormones [3,14]. The relative increase in serum alpha subunit is greater than that of serum TSH, resulting in a high molar ratio of serum alpha subunit to TSH (mean 3.2) [3,24].

The ratio is calculated using the following formula:

[Alpha subunit (in micrograms/L) ÷ TSH (in mU/L)] x 10

In postmenopausal women, who have normally elevated serum alpha-subunit levels accompanying their high serum gonadotropin levels, the cutoff which suggests a TSH-secreting pituitary adenoma is higher than in premenopausal and male patients (as high as 29.1 compared with 0.3 in men) [2,25]. The ratio can be normal in patients with microadenomas [14].

Imaging studies — The thyroid appearance on ultrasound and radioiodine imaging is similar to that in Graves' disease (diffuse homogeneous enlargement, normal or high radioiodine uptake, increased color Doppler flow).

Magnetic resonance imaging (MRI) of the pituitary more often shows a macroadenoma than a microadenoma, and some are locally invasive [14]. In one series of 18 patients, the mean macroadenoma size was 3.1 cm [26].

Indium-0111 octreotide scintigraphy may show focal uptake in the pituitary [27].

DIAGNOSIS — A TSH-secreting pituitary adenoma should be suspected in hyperthyroid patients with diffuse goiter and no extrathyroidal manifestations of Graves' disease, who have high serum free T4 and T3 concentrations and unsuppressed (normal or high) serum TSH concentrations, particularly in the presence of headache or clinical features of concomitant hypersecretion of other pituitary hormones (eg, symptoms of acromegaly). The subsequent finding of a pituitary macroadenoma by magnetic resonance imaging (MRI) is very strong evidence that the patient has a TSH-secreting pituitary adenoma, particularly in the presence of an elevated alpha subunit. (See 'Diagnostic evaluation' below.)

Differential diagnosis

Assay interference — Other conditions to be distinguished from TSH-secreting adenomas are those in which there is methodological interference in the measurement of total T4, total T3, or TSH. Serum total T4 and total T3 concentrations may be increased because of increased protein binding of the hormones in serum. These conditions include elevations in serum T4-binding globulin concentrations, familial dysalbuminemic hyperthyroxinemia (in which an abnormal albumin with increased affinity for T4 is produced), and the presence of anti-T4 antibodies. Patients with these conditions are euthyroid and have normal serum TSH concentrations, elevated total T4, and elevated total T3 but usually normal serum free T4 and free T3 concentrations when measured by appropriate methods. (See "Euthyroid hyperthyroxinemia and hypothyroxinemia".)

The presence of heterophilic antibodies can interfere with TSH measurements in immunometric assays. These human anti-mouse gamma globulins can bridge the two mouse monoclonal antibodies (solid phase antibody and signal antibody) and cause spuriously elevated readings for TSH [19,28]. Nonlinearity of TSH measurements with serial dilution of the patient's serum strongly suggests assay interference. Addition of nonimmune homologous mouse immunoglobulins has reduced this type of assay interference. Commercial assays exist for detecting human anti-mouse antibodies (HAMA).

In addition, autoantibodies to TSH have also been described which create TSH-anti-TSH immunoglobulin G (IgG) complexes (also called macro-TSH), which lack biologic activity but may be immunoreactive and cause spuriously high TSH values (often >100 mU/L) [29-31]. Autoantibodies to TSH can be detected by removal of the IgG-TSH complexes with polyethylene glycol or protein A or G, then repeating the assay on the immunosubtracted sera.

Patients with substances interfering with TSH measurements are usually euthyroid, not thyrotoxic, and have normal free T4 and T3. Thus, tests for interference with the TSH measurement are rarely required.

Resistance to thyroid hormone — Patients with TSH-secreting adenomas and hyperthyroidism must be distinguished from those with the syndrome of resistance to thyroid hormone due to mutations in the THRB gene (RTH-beta). Patients with RTH-beta have variable tissue hyporesponsiveness to thyroid hormone due to a defect in the THRB gene [24]. Despite their RTH in many tissues (including the pituitary gland), some patients with RTH-beta have symptoms and signs of hyperthyroidism, particularly tachycardia, hyperactivity, and hyperreflexia, and many have goiter. (See "Resistance to thyroid hormone and other defects in thyroid hormone action", section on 'Resistance to thyroid hormone beta (RTH-beta and nonTR-RTH)'.)

The following findings help to distinguish TSH-secreting adenomas from RTH-beta:

●A mutation in the THRB gene is present in patients with RTH-beta but usually not in patients with TSH-secreting adenomas. There is one reported patient with RTH-beta and a TSH-secreting adenoma [32]. There is also a reported patient with RTH-beta and an "incidental" pituitary adenoma not secreting TSH [33].

●The serum alpha-subunit concentration is normal in RTH-beta but, as noted above, often high in patients with TSH-secreting adenomas. (See 'Alpha subunit' above.)

●The serum sex hormone-binding globulin (SHBG) concentration is high in patients with TSH-secreting pituitary adenomas, whereas the values are normal in RTH-beta. This difference reflects the expected action of excess T4 and T3 on hepatic SHBG production in hyperthyroidism and resistance to the hormones' action in RTH-beta.

●The serum TSH concentration increases in response to thyrotropin-releasing hormone (TRH) in patients with RTH-beta but not in most patients with TSH-secreting adenomas.

●Patients with RTH-beta are more likely to have a fall in serum TSH concentrations in response to administered T3 (90 versus 12 to 25 percent) [3].

●Patients with RTH-beta are more likely to have a decrease in the velocity of TSH suppression two hours after subcutaneous injection of a short-term somatostatin analog [34].

The use of color flow Doppler sonography of the thyroid gland has also been shown to distinguish between patients with TSH-secreting pituitary adenomas and those with RTH-beta [35]. After a standard T3 suppression test, used for the diagnosis of RTH-beta (50 micrograms/day for days 1 to 3, followed by 100 micrograms/day for days 4 to 6, and then 200 micrograms/day for days 7 to 9) [24], the increased color flow Doppler sonography pattern and peak systolic velocity normalized in 8 of 10 patients with RTH-beta but not in any of the eight patients with a TSH-secreting pituitary adenoma.

Incidental pituitary adenomas may occur in patients with RTH-beta due to THRB gene mutations. There is no evidence to suggest that THRB gene mutations cause pituitary adenomas, but the relatively common occurrence of incidental pituitary adenomas, when they do occur in patients with THRB gene mutations, can make it difficult to distinguish one from the other.

Longstanding primary hypothyroidism — An occasional patient with longstanding primary hypothyroidism has sufficient hyperplasia of the thyrotroph cells to cause pituitary enlargement [36]. The hyperplasia may be mistaken for a pituitary tumor, but the patients have clinical manifestations of hypothyroidism, low serum T4 and T3 concentrations, and very high serum TSH concentrations. These patients should not be confused with patients with TSH-secreting adenomas. The enlarged pituitary in hypothyroidism shrinks with T4 therapy. (See "Disorders that cause hypothyroidism".)

Diagnostic evaluation — A TSH-secreting pituitary adenoma should be suspected in a hyperthyroid patient with high serum free T4 and T3 concentrations and unsuppressed (normal or high) serum TSH concentration. In patients with these thyroid function tests, we perform a combination of tests to confirm the diagnosis as there is no single test that will absolutely diagnose this condition:

Initial assessment:

●Ask the patient to question other family members regarding abnormal thyroid function tests. Take a careful history for other potential pituitary abnormalities (eg, does the patient have amenorrhea or a visual field defect?).

●Repeat the thyroid tests in another lab (TSH, free T4 and T3). Assay interference is frequently dependent on the commercial kit used for measuring the hormone. Another strategy is to ask the original lab to assay the serum TSH in serial dilution; authentic TSH will dilute in parallel with the serial dilutions (eg, a serum TSH of 10 mU/L will be 5 mU/L when the serum is diluted 1:2).

●Measure alpha subunit, as approximately 50 to 85 percent of patients with TSH-secreting pituitary adenomas have high serum concentrations. A normal level makes the diagnosis less likely but does not exclude it.

●Measure serum SHBG. A normal value is evidence that the patient is not hyperthyroid and argues for RTH-beta or spuriously elevated TSH values.

●Measure other pituitary hormones. Approximately 20 to 25 percent of the adenomas secrete one or more other pituitary hormones, predominantly growth hormone or prolactin [2]. Therefore, we measure insulin-like growth factor-1 (IGF-1) and prolactin in patients with suspected TSH-secreting pituitary adenomas. Additionally, the presence of other hormonal deficiencies suggests the possibility of a possible sellar mass (eg, measure follicle-stimulating hormone [FSH] in a postmenopausal woman, testosterone in a young man, etc).

Subsequent assessment:

●In the absence of any history or testing suggestive of a pituitary adenoma, given the difficulty in distinguishing patients with TSH-secreting tumors from RTH-beta, we perform an analysis for mutations in the THRB gene in patients with elevated free T4, T3, and nonsuppressed TSH and normal serum alpha subunit. In some countries, including the United States, this is now a rather widely available test. Although 15 percent of patients with RTH-beta do not have mutations in the TR-beta, the presence of a mutation rules out the diagnosis of a TSH-secreting tumor (see 'Resistance to thyroid hormone' above) and may prevent the possible confounding discovery of a pituitary incidentaloma.

●In the presence of history or testing suggestive of a pituitary adenoma, or after exclusion of RTH-beta, we obtain an MRI of the pituitary with gadolinium.

The presence of a macroadenoma on MRI is strong evidence of a TSH-secreting tumor, particularly in the presence of an elevated alpha subunit.

The presence of a microadenoma on MRI is not specific for a TSH-secreting tumor and can be seen as an incidental finding in 10 percent of normal individuals [37].

Conversely, small pituitary tumors can be missed by MRI or computed tomography (CT). In a few patients, radiolabeled pentetreotide has been useful in detecting the adenomas [38]. Cases of ectopic TSH-secreting adenomas have been reported in the nasopharynx [39,40] and in the vomerosphenoidal junction [41].

●Rarely, radiolabeled pentetreotide and inferior petrosal sinus sampling may be indicated when MRI of the pituitary shows a microadenoma or is normal. (See "Pituitary incidentalomas".)

TREATMENT — The treatment approach outlined below is primarily based upon case series and clinical experience [2].

Initial — The initial treatment of a TSH-secreting pituitary adenoma is medical therapy (with somatostatin analogs) to restore euthyroidism prior to surgery. Once euthyroid, transsphenoidal resection of the tumor is the most appropriate definitive therapy for patients with TSH-secreting pituitary adenomas. With this approach, most patients with microadenomas (≤10 mm) will be cured. Approximately half of patients with macroadenomas will require additional therapy for residual disease. (See 'Residual disease' below.)

Medical therapy — Medical therapy (typically somatostatin analogs) is used to restore euthyroidism prior to definitive treatment with transsphenoidal surgery. In addition, somatostatin analogs sometimes reduce the size of the tumor prior to resection. The role of somatostatin analogs as a primary treatment (instead of surgery) for TSH-secreting pituitary adenomas requires further investigation. However, there may be a role for somatostatin analogs as primary treatment for patients who refuse or are unable to undergo pituitary surgery. (See 'Primary treatment' below.)

Restore euthyroidism prior to neurosurgery — Patients should be euthyroid prior to neurosurgery. For medical therapy to restore euthyroidism, we suggest a somatostatin analog.

●We start first with a short-acting somatostatin analog in order to determine if the patient can tolerate the medication and if it is effective in lowering the TSH. A suitable initial dose of a short-acting somatostatin analog is 50 micrograms subcutaneously twice daily, increasing to three times daily and then to 100 micrograms three times daily, with additional increments of 50 micrograms per injection as needed; serum TSH and free T4 concentrations should be measured at two- to three-week intervals.

●If tolerated and efficacious, the medication can be switched to a long-acting somatostatin analog, which is administered as 20 mg intramuscular (IM) every four weeks.

Treatment with somatostatin analogs typically is continued for three to four months in order to normalize thyroid hormone levels. In a review of 43 cases of TSH-secreting pituitary adenomas, 26 patients received somatostatin analogs as initial therapy [14]. A reduction of more than 50 percent in TSH occurred in 23 of 26 patients (88 percent) and normalization of free T4 in 22 of 26 (85 percent). The side effects of somatostatin analogs include nausea, abdominal discomfort, bloating, diarrhea, glucose intolerance, and cholelithiasis. In addition, somatostatin analogs are very expensive. (See "Treatment of acromegaly".)

If somatostatin analogs are not tolerated, dopamine agonists (bromocriptine, cabergoline) may be effective in select cases, particularly in patients whose tumor cosecrete prolactin. However, in a series of 43 cases of TSH-secreting pituitary adenoma, seven patients were treated with dopamine agonists, and a significant reduction in TSH and prolactin occurred in only one of these cases, a patient with a mixed TSH/prolactin-secreting adenoma [14].

In addition to somatostatin analogs, a beta blocker such as propranolol (80 to 160 mg daily) or atenolol (25 to 50 mg daily) can be given to ameliorate some of the symptoms and signs of hyperthyroidism until more effective treatment is administered. (See "Beta blockers in the treatment of hyperthyroidism".)

Antithyroid therapy of any type, radioiodine or drug (eg, thionamide), is not indicated for the treatment of patients with TSH-secreting adenomas, because sustained reductions in thyroid hormone secretion would be expected to increase TSH secretion and stimulate tumor growth. However, if euthyroidism cannot be achieved with somatostatin analogs or dopamine agonists, short-term administration of a thionamide is necessary to restore euthyroidism prior to neurosurgery. (See "Nonthyroid surgery in the patient with thyroid disease", section on 'Hyperthyroidism'.)

Reduce tumor size prior to neurosurgery — Although the main goal of somatostatin analogs is to restore euthyroidism prior to surgery to minimize the risks of surgery, somatostatin analogs also may reduce the size of the adenoma prior to transsphenoidal surgery. As an example, in a review of 43 cases of TSH-secreting pituitary adenomas, 26 patients received somatostatin analogs as initial therapy [14]. Tumoral shrinkage of more than 20 percent occurred in 5 of 13 (36 percent) treated for at least three months. In another study, 23 of 38 (61 percent) of tumors shrank with octreotide after a mean of 33 days [42].

Primary treatment — The role of somatostatin analogs as a primary treatment for TSH-secreting pituitary adenomas requires further investigation. For patients receiving somatostatin analogs as primary therapy to control hyperthyroidism, thyroid function (TSH, free T4, total T3) should be monitored after two to three months, as somatostatin receptor analogs can induce TSH deficiency, necessitating a reduction in the frequency of injections [43].

In a retrospective study of seven patients (three with microadenoma, four with macroadenoma) who were treated with somatostatin analogs and followed for a mean of 8.5 years, six patients achieved good biochemical control [44]. Adenoma volume decreased in five of six patients for whom data were available. In another report, a patient with a 1.5 cm macroadenoma was treated with octreotide for four years, with complete resolution of hyperthyroidism and normalization of magnetic resonance imaging (MRI) findings [45]. After withdrawal of octreotide, the patient remained euthyroid and MRI scans have shown no recurrence during five years of follow-up. These findings suggest a primary role for medical therapy in some patients. However, in another case series, tumoral shrinkage was not observed in three of seven patients treated with octreotide for 24 months [14].

Why some tumors shrink in response to octreotide and others do not may be related to the expression of subtypes of somatostatin receptors (SSTRs) [46]. Octreotide may inhibit TSH secretion in all tumors that express SSTR2, and the expression of the type 5 receptor may enhance the inhibitory effects of octreotide [47].

Bromocriptine (10 to 20 mg/day orally) has proven effective as primary medical therapy in occasional patients, usually those with concomitant hyperprolactinemia, as has cabergoline, which is often better tolerated and need be given only once or twice weekly in a dose of 0.25 or 0.5 mg orally [48]. Occasional patients have responded to bromocriptine alone [49]. In three such patients, there was no growth of the pituitary tumor over eight years of therapy [50].

Surgery — Transsphenoidal resection of the pituitary adenoma is the definitive therapy of choice for patients with TSH-secreting adenomas. It results in cure in the majority of patients with microadenoma and approximately 50 to 60 percent of patients with macroadenoma [2,13,14]. In a review of 43 cases of TSH-secreting pituitary adenoma, transsphenoidal surgery was performed in 36 patients [14]. After one year, 21 (58.3 percent) were cured. The remaining 15 patients were treated with either radiotherapy (n = 8) or long-term somatostatin analogs. Surgical remission rates in micro- versus macroadenomas were 6 of 7 and 15 of 29 cases, respectively. (See "Transsphenoidal surgery for pituitary adenomas and other sellar masses".)

Criteria for cure — There are no well-established criteria for cure after transsphenoidal surgery [2]. The following parameters should be taken into account when assessing for cure:

●Clinical remission of symptoms of hyperthyroidism

●Normalization of thyroid function tests

●Absence of residual tumor on MRI

Somatostatin analogs and antithyroid drugs (if used) must be discontinued prior to the assessment of thyroid function tests. We discontinue somatostatin analogs three days prior to surgery and antithyroid drugs (thionamides, if used) on the day of surgery. We taper and discontinue beta blockers in the early postoperative period.

Thyroid function should be monitored carefully postoperatively. Assessment of thyroid function requires measurement of TSH, free T4, and total T3 at frequent intervals postoperatively, depending on the clinical scenario, until the patient's status is ascertained and steady-state conditions are established. While there are no published protocols, the first measurement of thyroid function tests should be 24 hours postoperatively, then weekly for four weeks, and then monthly for several months until the outcome of the surgery is certain:

●Normal TSH, free T4, and T3 – Successful surgery will restore euthyroidism. However, it may take months for thyroid function tests to return to normal. In one small series, the finding of an undetectable TSH level seven days after surgery was predictive of a successful outcome [51]. In another report, a TSH below 0.6 mU/L (obtained 12 hours postoperatively) was a major predictor of long-term remission [52].

●High TSH, free T4, and/or T3 – If the surgery is not successful, patients may develop recurrent hyperthyroidism after medical therapy is discontinued. However, if there was incomplete removal of the tumor, an apparent cure may be transient, and recurrent hyperthyroidism may occur months or years later with tumor regrowth.

●Low TSH and low free T4 – Surgery may damage the normal thyrotrophs (as well as other pituitary cells), resulting in central hypothyroidism or panhypopituitarism [53].

●Recovery of the pituitary-thyroid axis may take as long as two to three months, resulting in transient central hypothyroidism (low TSH and low free T4).

If the patient is started on thyroid hormone replacement for central hypothyroidism, especially in the absence of other pituitary deficiencies, slightly lower than full replacement doses should be given for the first few months to be certain that the central hypothyroidism is not transient. (See "Central hypothyroidism", section on 'Treatment'.)

Four to six weeks after discharge from the hospital, we also assess hormonal function of the nonadenomatous pituitary, regardless of whether it was normal or abnormal prior to surgery. We measure an early morning serum cortisol and either serum testosterone in males or serum estradiol and follicle-stimulating hormone (FSH) in premenopausal females. (See "Treatment of gonadotroph and other clinically nonfunctioning adenomas", section on 'First evaluation post-discharge' and "Diagnostic testing for hypopituitarism".)

An MRI should be obtained approximately 12 weeks postoperatively, and patients should be monitored for tumor recurrence based on clinical impression of tumor regrowth (headaches, visual changes) or recurrence of abnormal thyroid tests. (See 'Long-term monitoring' below.)

Residual disease — Because of the relatively poor results of surgery, many patients with macroadenomas need additional therapy [14]. For patients with residual disease after transsphenoidal surgery, we prefer to treat with somatostatin analogs rather than pituitary irradiation due to the more immediate response with medication [14]. However, for patients who cannot tolerate somatostatin analogs or who prefer to avoid long-term medical therapy, pituitary radiation is an alternative option to control TSH secretion in patients with persistent disease after transsphenoidal surgery [54].

The choice between long-term somatostatin analogs and pituitary radiation is based upon discussion with the individual patient about the risks and benefits of each therapy. The benefit of radiation therapy is the freedom from taking medications long term. However, adverse effects of pituitary radiation in the long term may include hypopituitarism, infertility, and, rarely, impaired cognitive function. Long-acting somatostatin analogs are injected monthly. Adverse effects include nausea, abdominal discomfort, bloating, diarrhea, glucose intolerance, and cholelithiasis. In addition, somatostatin analogs are very expensive.

Somatostatin analogs — Somatostatin analogs are administered after transsphenoidal surgery for long-term control of residual disease. As an example, in a series of 73 patients treated with octreotide (50 to 750 micrograms given subcutaneously two or three times daily), most of whom had already undergone surgery, the following results were noted [3]:

●The serum TSH concentration fell by more than 50 percent in 92 percent of patients and became normal in 79 percent.

●Serum T4 and T3 concentrations fell to normal in 95 percent of patients after one year.

●The tumor decreased in size in 52 percent of patients after one year.

●Less than 10 percent of patients developed resistance to the action of octreotide.

A slow-release formulation of a somatostatin analog, lanreotide, has proven effective in patients with TSH-secreting pituitary adenomas [55]. In 16 patients (the majority of whom had residual disease after transsphenoidal surgery), 30 mg lanreotide given IM two to three times per month for six months resulted in amelioration of hyperthyroid symptoms in all patients and normalization of serum TSH and thyroid hormone concentrations in 13, but no change in tumor size was observed. Thyroid tests (TSH, free T4, total T3) should be monitored as somatostatin analogs can induce TSH deficiency, necessitating a reduction in dosing frequency [43].

Pituitary irradiation — Radiation therapy may be effective in reducing the size of TSH-secreting adenomas and decreasing TSH, free T4, and T3 concentrations. There are few data on the efficacy of radiation therapy [13,14,54]. In one series, 22 patients had pituitary surgery for a TSH-secreting pituitary adenoma [13]. In the absence of surgical cure, 11 patients received external radiation of the pituitary (typical dose 4500 to 5500 cGy), and one patient received focused high-dose irradiation with a Gamma Knife unit. Patients received octreotide or bromocriptine until TSH secretion was controlled by radiation therapy as the full effect of radiation requires months to years. After 1 to 14 years of follow-up, the majority of the patients were biochemically euthyroid without requiring long-term treatment with somatostatin analogs, and 5 of 11 had no evidence of tumor on MRI.

Thyroidectomy — Thyroid surgery should be reserved for patients with symptomatic goiters in whom pharmacotherapy has failed. With the general availability of somatostatin analogs, however, thyroid surgery is rarely required. In patients with TSH-secreting pituitary tumors who undergo thyroidectomy either due to goiter or in whom pharmacotherapy or surgery of the pituitary lesion have failed, thyroid hormone replacement can be initiated at a dose which maintains the serum free T4 concentration in the upper 50 percent of the normal range. Serum TSH cannot be used to monitor therapy, since the TSH value may not be suppressible.

The risk of reactive growth of the pituitary cells has not been reported in patients with TSH-secreting pituitary tumors who receive normal replacement doses of thyroid hormone. This may reflect the relative nonaggressive nature of these tumors or the irrelevance of thyroid hormone feedback to tumor thyrotrophs. However, in those patients in whom tumor growth occurs despite appropriate T4 treatment, irradiation of the tumor is recommended.

Long-term monitoring — Patients with TSH-secreting adenomas who appear to be cured require monitoring of TSH, free T4, and free T3 two or three times in the first postoperative year and less frequently (annually) thereafter [2]. MRI of the pituitary should be performed one year postoperatively and then less frequently (every two to three years and then even less frequently) if the MRI is normal and there is no clinical or biochemical evidence of recurrence. For patients with mixed tumors (ie, TSH/growth hormone-secreting tumors), this includes no evidence of secretion of any components of the tumor. MRI should be repeated more often if there are any indications of biochemical or clinical recurrence.

Outcome — While most patients do reasonably well, in one series of 25 patients there were three deaths, including one from metastatic thyrotroph carcinoma [13]. An additional case of a metastatic TSH-secreting pituitary adenoma cosecreting prolactin has been reported [56]. In a series of 43 cases followed for a mean of eight years, two patients died (ages 75 and 86 years); 19 patients were cured after pituitary surgery alone; 17 patients with residual disease after surgery were controlled with somatostatin analogs, pituitary radiation, or both; and 7 were controlled by somatostatin analogs alone (no surgery) [14].

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Pituitary tumors and hypopituitarism".)

SUMMARY AND RECOMMENDATIONS

●Epidemiology – Thyroid-stimulating hormone (TSH)-secreting pituitary adenomas are a rare cause of hyperthyroidism, accounting for much less than 1 percent of all cases of hyperthyroidism. (See 'Introduction' above.)

●Clinical presentation – Most patients have the typical symptoms and signs of hyperthyroidism, but a few patients have mild or even no symptoms. Other clinical features include a diffuse goiter, visual field deficits, headache, and, in women, menstrual disturbances and galactorrhea. (See 'Clinical presentation' above.)

●Thyroid function tests and alpha subunit – The characteristic biochemical abnormalities in patients with hyperthyroidism caused by a TSH-secreting adenoma are normal or high serum TSH concentrations, high serum total and free thyroxine (T4) and triiodothyronine (T3) concentrations, and, in some cases, elevated serum concentrations of the alpha subunit of glycoprotein hormones. (See 'Thyroid function tests' above and 'Alpha subunit' above.)

●Differential diagnosis – Other conditions to be distinguished from TSH-secreting adenomas are those in which there is methodological interference in the measurement of total T4, total T3, or TSH and the syndrome of resistance to thyroid hormone beta (RTH-beta). (See 'Differential diagnosis' above and "Resistance to thyroid hormone and other defects in thyroid hormone action" and "Euthyroid hyperthyroxinemia and hypothyroxinemia".)

●Diagnostic evaluation – In patients with high serum free T4 and T3 concentrations and unsuppressed (normal or high) serum TSH concentration, we perform a combination of tests (eg, repeat thyroid tests in another lab, alpha subunit, other pituitary hormones) to confirm the diagnosis of TSH-secreting pituitary adenoma as there is no single test that will absolutely diagnose this condition. (See 'Diagnostic evaluation' above.)

•In the absence of any history or testing suggestive of a pituitary adenoma, given the difficulty in distinguishing patients with TSH-secreting tumors from RTH-beta, we perform (when available) an analysis for mutations in the thyroid hormone receptor beta (THRB) gene in patients with elevated free T4, T3, and nonsuppressed TSH and normal serum alpha subunit. Although 15 percent of patients with RTH-beta do not have mutations in the TR-beta, the presence of a mutation rules out the diagnosis of a TSH-secreting tumor. (See 'Diagnostic evaluation' above.)

•In the presence of history or testing suggestive of a pituitary adenoma, or after exclusion of RTH-beta, we obtain magnetic resonance imaging (MRI) of the pituitary with gadolinium. (See 'Diagnostic evaluation' above.)

●Diagnosis – The presence of high serum free T4 and T3 concentrations and measurable (normal or high) serum TSH concentrations in the presence of a macroadenoma on MRI is strong evidence of a TSH-secreting tumor, particularly in the presence of an elevated alpha subunit. (See 'Diagnosis' above.)

●Treatment

•Definitive therapy: Surgery – For definitive therapy of TSH-secreting pituitary adenomas, we recommend transsphenoidal resection of the tumor performed by a neurosurgeon with considerable experience (Grade 1B). The role of somatostatin analogs as a primary treatment for TSH-secreting pituitary adenomas requires further investigation. (See 'Treatment' above.)

•Medical therapy prior to surgery – Medical therapy is used to restore euthyroidism prior to surgery. For medical therapy, we suggest a long-acting somatostatin analog (Grade 2B). Although the main goal of somatostatin analogs is to restore euthyroidism prior to surgery to minimize the risks of surgery, somatostatin analogs also may reduce the size of the adenoma prior to transsphenoidal surgery. Dopamine agonists are an alternative option for those who do not tolerate somatostatin analogs, particularly in patients with cosecretion of prolactin. (See 'Restore euthyroidism prior to neurosurgery' above.)

A beta blocker such as propranolol (80 to 160 mg daily) or atenolol (25 to 50 mg daily) can be given to ameliorate some of the symptoms and signs of hyperthyroidism.

Long-term antithyroid drug therapy or thyroid ablation with radioiodine or surgery is not indicated in patients with TSH-secreting adenomas, because sustained reductions in thyroid hormone secretion would be expected to increase TSH secretion and stimulate tumor growth. However, if euthyroidism cannot be achieved with somatostatin analogs or dopamine agonists, short-term administration of a thionamide is necessary to restore euthyroidism prior to neurosurgery. (See 'Treatment' above and "Beta blockers in the treatment of hyperthyroidism" and "Nonthyroid surgery in the patient with thyroid disease", section on 'Hyperthyroidism'.)

●Criteria for cure – There are no well-established criteria for cure after transsphenoidal surgery. The parameters that should be taken into account when assessing for cure include clinical remission of symptoms of hyperthyroidism, normalization of thyroid function tests, and absence of residual tumor on MRI. (See 'Criteria for cure' above.)

●Residual disease – For patients with residual disease after transsphenoidal surgery, we prefer to treat with somatostatin analogs rather than pituitary irradiation due to the more immediate response with medication. The choice between long-term somatostatin analogs and pituitary radiation is based upon discussion with the individual patient about the risks and benefits of each therapy. (See 'Residual disease' above.)