Medical Management of Hormone-Secreting Pituitary Tumors

Pharmacologic Aspects

There are currently three DA available in the treatment of prolactinomas: cabergoline, bromocriptine, and quinagolide ( Table 14.1 ).

The semisynthetic ergot derivative, bromocriptine, a D2 agonist and weak D1 antagonist, was the first available medical treatment for prolactinomas. Administered orally, bromocriptine has a relatively short half-life (7 hours), often necessitating two to three daily doses, although once daily dosing may occasionally be effective. The standard dosage is 2.5 to 15 mg/day, and most patients are successfully treated with 7.5 mg or less. Since gastrointestinal side effects are common, it is usually recommended to start with very low doses (1.25 mg/day) and gradually titrate upward by 1.25 mg increments weekly until an effective dose is reached. In resistant patients, doses as high as 20 to 30 mg/day may be necessary.

Cabergoline, a preferential D2-receptor agonist, has largely supplanted bromocriptine as the first-line treatment of prolactinomas due to its superior efficacy, longer half-life, and greater tolerability. Its half-life of 2.5 to 4.5 days allows for once- or twice-weekly administration. The typical starting dose of cabergoline is 0.25 mg orally once or twice weekly, with escalation of dosing by 0.25-mg increments at 2- to 3-month intervals if plateau effect is reached, until PRL levels normalize. In most patients, the therapeutic dose range is 0.5 to 3.5 mg weekly. A study found that more than 95% of patients are able to achieve normal PRL levels with high-dose treatment (from 6 to 11 mg/week).

Quinagolide is a non-ergot derivative D ₂ -receptor agonist commonly used to treat prolactinomas in Europe and Canada. The half-life of quinagolide is 22 hours, which allows for a once-daily dosing regimen. As with cabergoline, the simpler dosing regimen and reduced side effect profile of quinagolide allows for better patient compliance compared with bromocriptine. Typical doses are 0.075 to 0.4 mg daily, and most patients are effectively treated with 0.1 mg/day. This drug can be rapidly titrated to therapeutic doses in just 7 days: a starting dose of 0.025 mg is given for 3 days and is titrated by 0.025 mg every 3 days until a dose of 0.075 mg is achieved.

Therapeutic Efficacy

Bromocriptine is 75% to 80% effective at normalizing PRL levels and shrinking tumor mass for microprolactinomas, whereas the success rate is around 70% for macroprolactinomas. ^{, ,} The improvement in headaches and visual field defects (mass effect of the pituitary tumor) is rapid and dramatic in most patients, occurring within a few days after the first dose. These changes often precede radiographic evidence of tumor shrinkage. Likewise, gonadal and sexual function may improve even before complete normalization of serum PRL levels, although sometimes normalization of testosterone may require several months of normal PRL levels. The degree of tumor shrinkage seen with bromocriptine does not necessarily correlate with the nadir PRL level or the extent of decline in PRL levels. Although most prolactinomas remain responsive to bromocriptine over time, the drug is generally not curative, as hyperprolactinemia and tumor growth will often recur after withdrawal of therapy. Prolonged use of bromocriptine has been associated with perivascular fibrosis and increased tumor consistency in prolactinomas that ultimately are surgically removed; the extent of fibrosis seems to correlate with the duration of presurgical treatment with bromocriptine and may affect the technical ease of surgical resection. ^,

Early studies found cabergoline to be 80% to 95% effective at normalizing PRL levels, with varying degrees of tumor reduction seen in 70% to 90% of patients. ^, Although there is considerable variability among studies with respect to tumor shrinkage, a large study showed a significant (>50%) tumor shrinkage in 80% of patients, and complete disappearance of the tumor mass in 57% of patients. Cabergoline is able to induce tumor shrinkage more effectively in DA naïve patients compared with those who have received prior DA treatment. Furthermore, cabergoline offers the potential for cure after several years of treatment, allowing for withdrawal of therapy. ^,

The beneficial effects of cabergoline on gonadal function are well documented in both sexes. Compared with bromocriptine, cabergoline is more effective at normalizing PRL levels and restoring ovulation and is associated with fewer, milder, and shorter-lived side effects. Serum testosterone levels have been shown to normalize in the majority of male patients, with concurrent improvement in sperm count and parameters.

Quinagolide normalized PRL levels in 70% to 100% of patients with microadenomas and 65% to 88% of patients with macroadenomas. ^, Significant tumor reduction was seen in 22% to 55% of patients with microadenomas and 25% to 75% of patients with macroadenomas. ^,

Tolerability and Side Effects

The most common side effects of DA include gastrointestinal, cardiovascular, and neurologic symptoms (see Table 14.1 ). In general, if a patient cannot tolerate the first DA administered, a trial of a second drug should be given.

Among the DA, bromocriptine is the least well-tolerated, with up to 12% of patients being unable to tolerate therapeutic doses. Up to one-third of patients will experience nausea and vomiting, an effect that can be minimized by initiating a low dose and uptitrating very slowly. Taking the medication with a small snack may also reduce these symptoms. A gradual dose titration and taking the medication at bedtime may also help reduce postural hypotension and dizziness experienced by as many as 25% of patients. Syncope is rare but has been reported even with small initial doses. Tolerance to postural hypotension usually develops rapidly. Drowsiness, headache, and nasal congestion are common complaints as well. The safety of bromocriptine in psychiatric patients has not yet been established; there are case reports of onset or exacerbation of preexisting psychosis. Other psychiatric symptoms associated with high doses of bromocriptine include anxiety, depression, insomnia, paranoia, and hyperactivity. At higher doses used in the treatment of Parkinson disease, reversible pleuro-pulmonary changes and retroperitoneal fibrosis have been reported, but these changes are unlikely to occur at the doses used to treat prolactinomas.

While cabergoline and bromocriptine have similar side effect profiles, cabergoline has been shown to be better tolerated in several large comparative studies. ^, Compared with bromocriptine, the side effects of cabergoline are generally less frequent, milder, and of shorter duration. Nausea and vomiting are most commonly observed, followed by headache and dizziness. In a multicenter randomized trial of hyperprolactinemic women, 3% could not tolerate cabergoline compared with 12% who had to stop taking bromocriptine. It has been proposed that fluctuations in concentrations of DA are the main cause of side effects; cabergoline would be better tolerated due to its long half-life and relatively steady plasma concentration.

Quinagolide has also shown better tolerability compared with bromocriptine. In a double-blind comparative study of 47 hyperprolactinemic patients, quinagolide was tolerated by 90% of patients versus 75% of patients treated with bromocriptine. As with the other DA, the most frequent side effects include nausea, vomiting, headache, and dizziness. These effects are transient and occur within the first few days of starting therapy or during dose adjustments.

Dopamine Agonists and Valvular Heart Disease

The safety of DA has been brought into question after two large population-based studies showed an increased risk of cardiac valvular disease in Parkinson disease patients being treated with high doses of these drugs, and particularly cabergoline and pergolide. ^, It is now recognized that one of the off-target effects of ergot-derived DA is activation of the serotonin (5-HT) receptor, of which there are seven distinct subtypes. High concentrations of the subtype 5-HT2B receptor are found on cardiac valves and the pulmonary arteries. Both pergolide and cabergoline have been implicated in the development of cardiac valvular fibrosis because of their agonist effects at the 5-HT2B receptor. Bromocriptine, which is only a partial agonist at the 5-HT2B receptor, had not been thought to pose an increased risk of cardiac valvular disease. However, some authors have questioned this and suggested that bromocriptine may, in fact, be equally associated with the development of cardiac valvulopathy at high cumulative doses. ^,

Several studies have examined the effect of cabergoline on cardiac valvular disease in patients with prolactinomas. These patients typically receive much lower doses (10 to 20 times less) than in Parkinson disease. One study reported an increased risk of moderate tricuspid regurgitation in patients who had received cumulative cabergoline doses greater than 280 mg. However, in another study, patients receiving cabergoline (0.25 to 4 mg/week) for 3 to 4 years (mean cumulative dose of 311 mg) had no increased prevalence of clinically significant valvular heart disease. Even in patients treated with cabergoline for 8 years who received cumulative doses as high as 1728 mg, there was no increased risk of clinically relevant heart disease. Other reports suggest that treatment with cabergoline in prolactinomas may correlate with the development of subclinical cardiac valvular disease, although a recent meta-analysis on the use of DA in patients with hormone-secreting pituitary tumors (prolactinomas, acromegaly, and CD) did not confirm this association.

In conclusion, conventional doses of cabergoline and bromocriptine used in prolactinomas may be associated to a slightly increased prevalence of clinically insignificant cardiac valve fibrosis and insufficiency. ^, However, it may be prudent to monitor patients requiring higher doses of ergot-derived dopamine agonists with serial echocardiograms (e.g., cabergoline > 2 to 3 mg/week, or bromocriptine > 10 mg/daily). If valvulopathy develops, it may be reasonable to switch to quinagolide, which is non-ergot derived and has low affinity for the 5-HT2BR. If quinagolide is not available, a reasonable option would be to switch from cabergoline to bromocriptine, which has a lower affinity for the 5-HT2B receptor. Alternatively, surgical resection may be appropriate. This issue underscores the importance of using the smallest effective dose and attempting withdrawal from cabergoline treatment in patients who achieve normalization of PRL levels and significant tumor shrinkage to minimize their cumulative lifetime exposure to the drug.

Dopamine Agonist Resistance

A subset of patients with PRL-secreting pituitary tumors demonstrate “biochemical” resistance (failure to normalize PRL levels) or “mass” resistance (absence of tumor shrinkage) to DA. Resistance to DA occurs in both micro- and macroadenomas and is believed to be mediated by the down-regulation of pituitary D2 receptors. Biochemical resistance has been estimated to occur in approximately 10% to 20% of patients on DA. Mass resistance has been estimated to occur in approximately 30% to 40% and 15% of patients taking bromocriptine and cabergoline, respectively. ^, The prevalence of resistance to quinagolide is unclear given the lack of data regarding its use in DA-naïve patients.

Most tumors resistant to bromocriptine will respond to cabergoline. When switched to cabergoline, 85% of patients resistant to bromocriptine and quinagolide achieved normal PRL levels, and 70% had some change in tumor size. A large study showed that cabergoline normalized PRL levels in 61% and 50% of bromocriptine-resistant microprolactinomas and macroprolactinomas, respectively.

Therapeutic options in the management of DA-resistant prolactinomas include (1) increasing the dose of the DA, (2) switching to an alternative DA, (3) trans-sphenoidal surgery, and (4) radiation therapy (RT). In most cases, the resistance to DA is partial, and a response can be achieved by progressively increasing the dose of the medication. A study showed that in more than 95% of patients resistant to bromocriptine PRL normalized with individualized high-dose cabergoline treatment (mean dose 5.2 mg/week). Bromocriptine-resistant patients generally require higher doses of cabergoline and have less tumor shrinkage compared to treatment-naïve patients. In general, switching from cabergoline to bromocriptine is unlikely to be effective.

While trans-sphenoidal surgery is almost never a first-line treatment for macroprolactinomas (see below), it remains an important option for patients who cannot tolerate DA or when medical therapy is ineffective at controlling symptoms or restoring reproductive function. There is little long-term outcome data regarding patients with macroprolactinomas treated with DA who subsequently proceed to surgery. In a study of 72 patients, 35% of subjects required trans-sphenoidal surgery due to resistance and/or intolerance to DA. This study, which is an outlier given the disproportionately high number of DA-resistant patients, showed that additional tumor shrinkage was achieved in 57% of operated patients, while only 22% were able to attain normoprolactinemia without DA following surgery. Surgery was associated with a high incidence of hypopituitarism regardless of whether patients received subsequent RT.

The remission rate for surgery when used as second-line treatment for macroprolactinomas is quite low (around 20%). Surgery often plays only a “debulking” role, and therefore should be reserved for patients who, despite maximal medical therapy, have progressive tumor enlargement and are at risk of visual compromise or neurological deficits.

Conventional fractionated RT should be considered a last-line option in the management of DA-resistant prolactinomas given the delayed treatment effects and the high rate of hypopituitarism.

Dopamine Agonist Withdrawal

The optimal duration of therapy for patients with prolactinomas is uncertain. Until recently, DA treatment had been considered a “lifelong” requirement. However, after a landmark study by Colao et al. demonstrated disease remission in a considerable proportion of patients following withdrawal from cabergoline, attention shifted toward defining selection criteria for withdrawal and identifying predictors of long-term remission. ^, Defining the optimal timing and appropriate candidates for withdrawal is difficult because of the heterogeneity of studies that have examined this issue, which differ with respect to the causes of hyperprolactinemia (idiopathic vs. micro/macroprolactinoma), the type and duration of DA treatment, and the treatment prior to the start of DA therapy. Despite the inherent difficulty in extrapolating from these disparate studies, criteria for withdrawal include (1) normoprolactinemia and (2) tumor absence or markedly reduced tumor volume after a minimum of 2 years of DA treatment. A study tested the applicability of these criteria and found a recurrence rate of 54% after withdrawal of long-term cabergoline treatment. In this study, most patients had recurrence of disease within 1 year of discontinuation, and a similar recurrence rate (52% vs. 55%) was seen in patients with microprolactinomas and macroprolactinomas. The size of the tumor remnant prior to withdrawal appears to be predictive of recurrence risk. A meta-analysis showed that long-term remission occurs in only 21% of patients following DA withdrawal.

A retrospective analysis of 213 patients with prolactinomas evaluated which factors predict a satisfactory response to DA treatment and the recurrence rate after discontinuing DA using the above criteria. Thirty-seven percent of patients were able to discontinue DA treatment; those with smaller tumor diameter, no extrasellar extension, and rapid reduction of PRL levels after initiating DAs were more likely to achieve withdrawal. However, hyperprolactinemia recurred in 75% of cases, with most cases recurring within 5 years. Patients who did not have cavernous sinus invasion at presentation and those with more pronounced PRL reduction immediately after starting DA were more likely to maintain long-term remission. These results emphasize that recurrence of hyperprolactinemia is common after discontinuing DA and patients with prolactinomas require long-term follow-up. Moreover, patients who are likely to need life-long DA treatment could be identified soon after diagnosis. This could allow for tailored management, e.g. considering surgery early in the management of patients with DA-related cardiac abnormalities, psychiatric symptoms, and behavioral changes.

The best chance for a long-term remission is seen in patients treated with cabergoline for more than 2 years, whereas remission is seen in only 20% of patients withdrawn from bromocriptine. Considering the financial burden, occasional problems with tolerability, and potential risk of valvulopathy with long-term DA treatment, it is reasonable to attempt withdrawal in patients who have been on a DA (particularly if cabergoline) for 2 years or more. Kwancharoen et al. tried to stop cabergoline a second time in a patient who previously failed withdrawal. In those who received an additional 2-year course of low-dose cabergoline (≤1 mg/week), persistent normoprolactinemia was observed in 35% of patients after withholding the medication.

Surgery as a First-Line Option in Microprolactinomas

Medical treatment with DA is the mainstay in the primary management of macroprolactinomas and symptomatic microprolactinomas. However, surgery may be considered as a first-line treatment in selected patients, including:

• 1.

  Patients taking antipsychotic medications: DA can cause exacerbation of the preexisting psychosis, and if the medical treatment is started the patient has to be followed up closely by the psychiatry service as well;

• 2.

  Women with large macroprolactinomas wishing to become pregnant. In fact, these patients (even if they are likely to respond to DA) would have to withhold medical treatment after conception and the adenoma may increase in size during pregnancy;

• 3.

  Patients with symptomatic hyperprolactinemia not willing to receive long-term medical treatment.

The latter scenario is gaining increasing attention in recent years, especially in terms of surgical approach to symptomatic microprolactinomas in patients with a long life expectancy. In fact, while macroprolactinomas are rarely surgically curable, trans-sphenoidal surgery can provide a safe, effective, and definitive cure in patients with microprolactinomas. On balance, successful surgery can avoid long-term medical treatment with DA, which are potentially associated with adverse events, albeit very rare. Trans-sphenoidal surgery for microprolactinomas leads to normal PRL in 71% to 100% of patients, with low rates of postoperative hypogonadism and permanent diabetes insipidus (0% to 6%) and minor neurosurgical complications (0% to 2%). The long-term recurrence of hyperprolactinemia in patients with normal postoperative PRL is around 10%. A recent cost analysis suggested that surgery may be a more cost-effective treatment for prolactinomas than medical management, although the cost difference was not impressive.

In conclusion, initial surgery can be an option in selected patients. The risks and benefits of surgery should be discussed extensively with the patient, provided that an experienced neurosurgeon is available. ^,