Diseases of the hypothalamusv(link to Von Werder chapter), such as tumors, arterio-venous malformations, and inflammatory processes such as sarcoidosis result in either diminished synthesis or release of dopamine. Furthermore, certain drugs (e.g. alpha-methyldopa and reserpine) are capable of depleting the central dopamine stores.

Section of the pituitary stalk results in impaired transport of dopamine from the hypothalamus to the lactotrophs. Pituitary or stalk tumors with abnormal blood supplies, or their pressure effects, may interfere with the circulatory pathway from the hypothalamus down the pituitary stalk to the normal lactotrophs or a tumor, producing effective dopamine deficiency due to a functional stalk section.

Dopamine receptors have been found on human pituitary lactotroph adenoma cells. Receptor sensitivity to dopamine may be diminished, which would explain the lack of response to increased endogenous dopamine stimulation; however, an obvious response of the receptors to pharmacologic dopamine agonists makes this possibility less likely. Certain drugs act as dopamine-receptor-blocking agents, including phenothiazines (e.g. chlorpromazine), butyrophenones (haloperidol), and benzamides (metoclopramide, sulpiride, and domperidone). These drugs block the effects of endogenous dopamine and thus release lactotrophs from their hypothalamic inhibition. This sequence of events results in hyperprolactinemia.

Hypothyroidism may be associated with hyperprolactinemia. If hypothyroidism results in increased TRH production, then TRH (which can act as a PRF) could lead to hyperprolactinemia. Estrogens act directly at the pituitary level, causing stimulation of lactotrophs, and thus enhance prolactin secretion. Furthermore, estrogens increase the mitotic activity of lactotrophs, increasing cell numbers. Injury to the chest wall can also lead to hyperprolactinemia; this results from abnormal stimulation of the reflex associated with the rise in prolactin that is seen normally in lactating women during suckling.

The first dopamine agonist ergot compound to be used in clinical practice was bromocriptine, a peptide ergot. It was introduced in the early 1970s in Europe and thus there is more than 30 years of experience of the use of such compounds in the treatment of hyperprolactinemia. Bromocriptine has the advantage of having a long duration of action compared to dopamine itself or oral compounds such a levo-dopa.

Bromocriptine has a similar mode of action to dopamine in stimulating dopamine receptors on the prolactin-secreting pituitary cells - D2 receptors. Stimulation of these receptors leads to inhibition of both prolactin secretion and synthesis. After a single 2.5mg dose of bromocriptine administered at 09.00h to women with hyperprolactinemia, prolactin secretion was inhibited within 2 hours and reached a nadir at 7 hours. When patients are treated chronically using 2.5mg three times per day, prolactin levels are maintained within the normal range, i.e. less than 20ng/mL (400mU/L), throughout a 24-hour period. Subsequently a variety of other compounds have been developed which are useful additions. These include pergolide mesylate, which only needs to be administered once per day, quinagolide, and cabergoline.

Cabergoline has an extremely long biological half life and thus only needs to be administered either once or twice per week at a dose of 0.5 mg/dose. In refractory cases the dose can be increased gradually to 1 mg up to a daily dose. In addition to its long biological half-life, cabergoline is generally better tolerated than bromocriptine, so increasing the patient's adherence. Therefore, cabergoline is currently considered the first choice drug for the treatment of prolactinomas, except for patients wishing pregnancy in the short-term (see below). A recent study showing that patients with Parkinson's disease on cabergoline or pergolide presented high prevalence of valvular hearth disease brought some concern about the drug safety. It must be pointed out, however, that the doses of dopamine agonists for Parkinson's disease are much higher than for prolactinomas.

The first male patient commenced bromocriptine treatment at St Bartholomew's Hospital, London, in 1971, and his case illustrates a number of important points. Initially, the prolactin levels of the patient were extremely elevated, but the administration of bromocriptine lowered them into the normal range (undetectable by bioassay); this was associated with cessation of galactorrhea. Bromocriptine therapy normalized the patient's prolactin levels and gonadal function, restoring potency. Also cabergoline showed to be very effective in treating prolactinoma in men, even those bearing giant adenomas. Interestingly enough, testostesterone administration to men without correction of hyperprolactinemia generally do not improve libido and potency.

From experience of treating a large number of amenorrheic hyperprolactinemia women, the results of treating the first 58 appear to be representative of the success that can be achieved with bromocriptine (and other dopamine agonists) therapy (Fig. 1). Within 1 month of starting therapy, a regular menstrual cycle is restored in approximately 25% of patients. Within 2 months, regular menstrual cycling can be restored in over 60% and, within 10 months, in some 80%. Those patients who did not experience restoration of regular menstrual cycles (with only one or two exceptions) had previously undergone surgery or radiotherapy for their pituitary tumor - this rendered them gonadotropin-deficient. Thus, if hyperprolactinemia is the cause of amenorrhea, the chances of restoring normal gonadal function by medical therapy alone are extremely good, around 85%.

Figure 1. Success rate of bromocriptine in amenorrhea. If hyperprolactinemia is the cause of amenorrhea, the chances of restoring normal gonadal function with bromocriptine are very good. After 1 month of treatment, one woman in four will return to normal menstrual cycling; within 2 months, this number will increase to six out of 10; and after 10 months, eight out of 10 women will be menstruating normally. (Most of the remaining 20% have had pituitary surgery and irradiation therapy and are gonadotropin-deficient.)

To study the long-term effects of bromocriptine on prolactin secretion, a group of patients was carefully evaluated (Fig. 2). Ten blood samples were taken from each patient before treatment; at 3, and 6 months on therapy. All patients were treated with the same dose of bromocriptine (2.5mg three times per day). In all cases, bromocriptine lowered prolactin levels.

Figure 2. Long-term effects of bromocriptine therapy. Prolactin levels in a group of patients on long-term bromocriptine therapy were tested before therapy; after 3 months, 6 months (during which time all patients received bromocriptine 2.5mg three times per day) In all cases, prolactin remained suppressed throughout the year and, in most cases, prolactin levels were held within the normal range. Gonadal function was restored, even in patients whose prolactin levels did not return to normal. Reproduced with permission from J Clin Endocrinol Metab 50:1026 1033.

Ovulation is normally associated with a dip in basal body temperature, and normal luteal function with a temperature rise. Basal body temperature is therefore a useful means of documenting ovulation. When hyperprolactinemic patients have had their prolactin levels and periods restored to normal by bromocriptine therapy, they usually demonstrate a biphasic temperature pattern. One patient had suffered from polymenorrhea for many years and was found to be hyperprolactinemic. Bromocriptine normalized her periods, and therapy was withdrawn after 1 year. During therapy, the basal body temperature chart showed a normal biphasic pattern but, following withdrawal of bromocriptine, prolactin levels rose to more than 100ng/mL (2000mU/L), galactorrhea returned, and the temperature pattern immediately became monophasic. Although the patient did not become amenorrheic, she developed irregular (presumably anovulatory) periods. Two weeks following reinstitution of therapy, the patient ovulated, demonstrating a postovulatory temperature rise. She has subsequently had a regular cycle and three successful pregnancies, each with the help of bromocriptine.

The pathogenesis of the hypogonadal state in hyperprolactinemia is poorly understood. In men, testosterone levels are usually low but can occasionally be normal, while in women, a hypoestrogenic state may occur, with loss of ovulation. The clinical features in hyperprolactinemic women, however, differ from those in the postmenopausal state since breast atrophy is absent and gonadotropin levels are not elevated.

The mechanisms which may account for the suppression of gonadal function in hyperprolactinema include suppression of gonadotropin secretion; inhibition of positive estrogen feedback on luteinizing hormone (LH) secretion in women; an increase in adrenal androgen secretion; and blockade of the effects of gonadotropins at the gonadal level. It is probable that an important mechanism is prolactin feedback at the hypothalamus, which alters secretion of gonadotropin-releasing hormone (GnRH), thus causing LH and follicle-stimulating hormone (FSH) secretion to become inappropriately low relative to gonadal steroid levels. Reduction in the normal LH pulsatility, essential for normal gonadal function, also occurs. Prolactin may interfere with LH and FSH action at the gonad, blocking progesterone synthesis, and may stimulate adrenal androgen secretion.

Surgical therapy of large prolactin-secreting pituitary tumors is unsatisfactory since it is only capable of normalizing serum prolactin levels or gonadal function in fewer than 20% of patients, particularly those with high prolactin levels, so there is a need for another approach to the problem. Three major lines of evidence suggest that medical therapy may help in the treatment of these large tumors.

Bromocriptine reduces pituitary tumor size in 75-80% of patients with large prolactin-secreting tumors, even with gross extrasellar extension. The type of result that can be expected is illustrated by a patient with a large prolactin-secreting tumor, who was treated with bromocriptine alone. The patient had a suprasellar extension and visual field defects. Visual field plots from this patient before and during treatment, as well as after withdrawal and reinstitution of bromocriptine therapy, are illustrated in Figure 3. Before therapy (baseline), the patient had a bitemporal hemianopia, complete in the left eye and incomplete in the right eye. The visual fields were greatly improved after 10 days, and only an equivocal superior bitemporal quadrantic defect to the low intensity object was present after nearly 1 year. By 13 days after withdrawal of medical treatment, the tumor had enlarged again and the field defects recurred as an almost complete temporal hemianopia in the left eye and a partial temporal hemianopia in the right. Progressive improvement in visual fields was again observed over the subsequent 6 months after reintroduction of therapy.

Figure 3. Visual field plots in hyperprolactinemia. The visual fields shown here were plotted using a Goldmann perimeter, under identical conditions, with a 0.25mm2 object at two different light intensities: 1000 apostilb (I4) and 100 apostilb (I2). The black periphery indicates a normal visual field for comparison. An almost complete bitemporal hemianopia (pretherapy), which had almost disappeared after 1 year of treatment with bromocriptine, returned on cessation of therapy and began to subside after reinstitution of bromocriptine. Reproduced with permission from J Clin Endocrinol Metab 53:480 483

Figure 4 illustrates coronal CT head scans (postenhancement) from the patient whose visual fields are shown in Figure 4(a), performed on a Delta 25 scanner, illustrates the situation before therapy; Figure 4(b) shows a scan performed on a GE 8800 scanner 2 weeks after starting bromocriptine therapy, 2.5mg three times per day. Before therapy, Figure 4(a) shows an enlargement of the pituitary fossa and an enhancing mass extending inferiorly into the sphenoid sinus, superiorly into the chiasmatic cistern, and abutting on the third ventricle. Two weeks post-treatment, Fig. 4(b) shows a marked reduction in tumor size, with regression of the suprasellar extension. The chiasmatic cistern is now largely free of tumor, apart from a finger-like process to the left of the midline. The intrasellar high density is present in the pre-enhancement scan and represents calcification within the tumor. Within the short space of 2 weeks, therefore, there was marked reduction in the size of the pituitary and a consequent decompression of the optic chiasm, which explains the rapid improvement in visual fields observed in this patient.

Figure 4. Coronal CT head scans before and during treatment with bromocriptine. Magnetic resonance imaging  (T1 weighted with gadolinium enhancement) of a man harboring a large macroprolactinoma. A1(coronal view) A2 (sagittal view) before treatment and B1 (coronal view) and B2 (sagittal view), after one month of cabergoline (0,5mg thrice a week) treatment. The significant tumor shrinkage was paralleled by serum prolactin decrease from 6,600  ng/mL to 540 ng/mL and by visual improvement

Of 16 patients with large prolactin-secreting tumors, 13 showed similar changes. The three patients in whom these changes were not observed were:

Other groups have had similar results with other dopamine agonist drugs. It seems that at least 65% of macroadenomas with large extrasellar extensions (or even more with cabergoline) may be treated with dopamine agonists alone to shrink the tumors and to relieve the mass effects and the hormonal excess.

In patients with macroadenomas, the serum prolactin levels can be readily suppressed with bromocriptine therapy. Figure 5 shows serum prolactin levels throughout the day after an initial 2.5mg oral dose of bromocriptine administered at 09.00h to the patient whose visual field plots and CT scans are shown in Figures 3 and 4, as well as those from a patient with a similar problem. After a single dose of bromocriptine, the prolactin levels fell by approximately 90%. The mean and absolute range of prolactin levels in samples taken at the same time intervals before therapy (baseline) and during bromocriptine therapy, 7.5mg/day, are also shown in Figure 6. In the first patient, the prolactin levels were suppressed into the normal range (less than 18ng/mL, 360mU/L) and in the second patient, prolactin levels, although lowered, did not return to normal. With treatment over 1 year, however, the levels continued to fall to 78ng/mL (1560mU/L). In these patients, as with the patients with microadenomas, gonadal function is usually restored to normal. As previously noted, when therapy was withdrawn at 1 year, visual field defects recurred in the first patient and this was associated with prolactin levels rising again to 2580ng/mL (51.6u/L) at 13 days in comparison to the pretreatment level (3940ng/mL, 78.8 u/L). However, it should be noted that, in male patients on bromocriptine, prolactin levels usually fall rapidly and easily into the normal range. If this does not occur, gonadal function may not return to normal.

Figure 5. Changes in serum prolactin levels after bromocriptine administration. The effect on serum prolactin levels throughout the day of a single 2.5mg oral dose of bromocriptine at 09.00h is shown. Case 1 is the patient whose visual field chart and CT scans are shown in Figures 3 & 4. Case 2 is a patient with a similar problem. In patients such as these, even when prolactin levels do not come down to the normal range, gonadal function is usually restored.Reproduced with permission from J Clin Endocrinol Metab 51:438 445.

Side effects of bromocriptine therapy occur only at the start of treatment and disappear with continued therapy. There are no long-term problems associated with chronic treatment at the doses used for hyperprolactinemia - usually 7.5mg/day and rarely more than 15mg/day.

If treatment is started with full doses or increased too quickly, dizziness, nausea, and postural hypotension may occur. To avoid such effects, bromocriptine must always be taken during a meal. Administration should be started at night, with a sandwich and glass of milk, when the patient retires to bed. After taking half a tablet (1.25mg) for three nights, half a tablet is added with breakfast. After a further three days an additional half a tablet is added with lunch. At intervals of three days, additional half tablets (1.25mg increments) may be progressively added until achievement of the usual dose of one tablet (2.5mg) taken three times daily, in the middle of breakfast, lunch, and the evening meal. If side effects still occur, longer intervals and smaller increments should be used. Once established on an effective dose it is now established that the whole dose can be given once daily, again during the main course of a main meal. In a small proportion of patients psychosis or anxiety can be precipitated by the administration of dopamine-agonist drugs. These are usually adequately managed either by stopping the dopamine-agonist medication or lessening the dose. These symptoms are more common in parkinsonian patients, but in this group the brain is diseased and the doses used are much greater than for the indication of prolactinoma.

The only group of patients who do not suffer from such side effects if given the full dose immediately is puerperal women. They may be given bromocriptine 2.5mg two or three times daily to suppress puerperal lactation, without side effects, if treatment is started within 24 hours of delivery. The reasons for this difference are unknown.

There are several other dopamine agonists that lower serum prolactin levels and reduce tumor size to a similar extent to bromocriptine. These drugs include pergolide, lisuride, qinagolide, and cabergoline. These compounds are associated with a similar side effect profile to that observed with bromocriptine. Cabergoline has the advantage that it only needs to be taken once or twice per week and may have a reduced incidence of side effects. With the exception of bromocriptine, safety during pregnancy has not been demonstrated. As experience with cabergoline has accumulated it appears that some patients who could not be controlled with other dopamine agonists, e.g. bromocriptine or pergolide, could be controlled by cabergoline. On occasion, the dosing may need to be increased as frequently as daily. Because of its long duration of action, the side effect profile and efficacy appear to be better than with any other dopamine agonist drug.

Can dopamine agonist drug be withdrawal without recurrence of hyperprolactinemia?

One of the drawbacks of medical treatment of prolactinomas is the need for long-term therapy. As a matter of fact, treatment with bromocriptine and other dopamine agonist drugs generally is considered as "symptomatic", since bromocriptine discontinuation leads to recurrence of hyperprolactinemia in most patients and to tumor regrowth, at least after short-term use. Concerning long-term therapy with bromocriptine, a recent retrospective study showed that 25.8% of 62 patients with microprolactinomas and 15.9% of 69 patients with macroprolactinomas treated with bromocriptine for a median time of 47 months persisted normoprolactinemic after a median time of 44 months of drug withdrawal. There were no statistically significant differences regarding age, gender, bromocriptine initial dose and length of use, tumor size, pregnancy during treatment, and previous pituitary surgery or radiotherapy among patients who persisted with normoprolactinemia and those who did not. Another study encompassed a large cohort of hyperprolactinemic patients on cabergoline. The drug was discontinued in patients who attained normoprolactinemia, at least 50% of tumor reduction or disappearance on image, and at least 2 years of follow-up after cabergoline withdrawal. It was demonstrated that serum prolactin remained normal in 76%, 70% and 64% of patients with "idiopathic" hyperprolactinemia, microprolactinomas and macroprolactinomas, respectively. This great discrepancy between results with bromocriptine and cabergoline was not confirmed by a recent study dealing just with microprolactinomas. The question regarding why long-term findings differ from short-term ones may be answered by the formerly described microscopic alterations of the lactotroph during bromocriptine administration, suggesting a cytostatic effect related to short-term therapy and a cytocidal one to long-term treatment, which could explain the maintenance of normoprolactinemia after drug withdrawal. Another factor that may influence remission of prolactinomas is their natural history. Nevertheless, the influence of the natural history of non treated prolactinomas, many of them reaching "spontaneous normoprolactinemia, cannot be rulled-out.

Many hyperprolactinemic women would like to become pregnant and, since the administration of dopamine agonist drugs lowers the prolactin levels and restores gonadal function, conception resents little difficulty. There are, however, several important considerations that must be recognized by both physician and patient, including the possible teratogenic sequelae of fetal exposure to bromocriptine and other drugs.

There is no evidence for teratogenicity in animal studies and, in 1400 women who were taking bromocriptine when they conceived, there is no evidence of increased incidence of abortion, multiple pregnancy, or fetal abnormalities. Until these babies have lived their own complete life cycles, however, the possibility of unexpected late effects cannot be excluded. In order to minimize fetal exposure to bromocriptine, it is suggested that patients should initially use mechanical contraception. Once regular menstrual cycles have occurred, preferably at least three, contraceptive precautions are discontinued. In this way, pregnancy can be suspected as soon as a menstrual period is 48 hours overdue. At that time, a serum human chorionic gonadotropin assay should be performed to confirm the pregnancy, and the patient should discontinue bromocriptine therapy. In this way, the fetus is exposed to bromocriptine for a theoretical maximum of 16 days.

There is little doubt that patients with pituitary tumors run a small, but significant, risk of expansion of the tumor during pregnancy. It is very difficult, however, to assess the absolute risk. With microadenomas which did not undergone prevous surgery or radiotherapy, the incidence seems to be between 1,6% and 5,5%.In patients with macroadenomas also not operated on or irradiated , the incidence is higher, between 15.5% and 41%. However, the risk of complications may be lower in women previously operated or irradiated. This risk is unrelated to bromocriptine therapy prior to pregnancy but may occur when fertility is induced with other drugs, including exogenous gonadotropins and clomiphene, and even when no drug therapy has been employed in patients with pre-existing pituitary adenomas.

In practice, the problem of pregnancy is not great, since the vast majority of women who present with hyperprolactinemia only have microadenomas. To avoid major problems, it is extremely important that patients undergo careful endocrine, neuroradiologic, and neuro-ophthalmologic evaluation prior to treatment. If there is no suprasellar extension, and if the patient harbors only a microadenoma, then the risk of clinically significant swelling of the pituitary is extremely small; it is therefore suggested that the patient is evaluated clinically at bimonthly intervals throughout the pregnancy. If the patient has a macroadenoma and suprasellar extension, a strong case can be made for transsphenoidal decompression of the tumor prior to pregnancy. However, in those patients with a good response to dopamine agonists in terms of prolactin normalization and tumor shrinkage within sellar boundaries, at least one year before pregnancy, the drug can be withdrawal and reintroduced if tumor re-growth is observed If such approach fails, pituitary surgery or premature delivery, if feasible, would be indicated. Additionally, in case of tumor apoplexy high-dose dexamethasone may improve clinical symptoms and also may reduce the chances of fetal respiratory distress if premature delivery is needed.

In recent years, pregnancies have been described in patients using other DA such as quinagolide, and cabergoline. Although there is no evidence of increased frequency of abortions or malformations in cabergoline-induced pregnancies, the drug's long action, which persists up to three weeks after its withdrawal, associated with fewer data when compared to bromocriptine (around 400 versus over 6,000 pregnancies), limit the confidence that can be used in prescribing it patients who wish to conceive or its use during pregnancy. In the Uniited States bromocriptine is the only FDA approved drug for inducing pregnancy in hyperprolactinemic women.

It is, however, possible even for these patients to go through pregnancy without developing visual disturbances and, furthermore, even if visual disturbances occur in one pregnancy, the problem may not recur in subsequent ones.

Thus, the approach to the patient with a prolactin-secreting macroadenoma who desires pregnancy can be either expectant or prophylactic. The authors believe that, as the risk of swelling of the adenoma is less than 20%, it is reasonable to adopt an expectant policy. Others suggest that pituitary decompression should be performed surgically, and still others recommend that external pituitary irradiation is given. It is not clear whether external pituitary irradiation or decompression of the pituitary by surgery, or both, completely prevent symptom-generating pituitary enlargement. It should be stressed that, so far, no patient has become permanently blind following expansion of the tumor during pregnancy.

If visual field defects or headaches from tumor expansion do occur during pregnancy, a number of therapeutic options are available. Following termination of the pregnancy, either by abortion or delivery of the baby, tumors have become smaller and such symptoms and headaches have resolved in all cases. Thus, if such symptoms occur early in pregnancy, therapeutic termination may be indicated. If they occur in the eighth month of pregnancy, premature delivery of the baby may be decided upon, although, if field defects and symptoms are minor, careful observation may be all that is required. The most problematic situation arises when symptoms occur in the middle trimester. At that time, it is suggested that bromocriptine therapy is restarted and in the great majority of patients there is rapid reduction in the tumor size and further swelling is prevented. If this is unsuccessful, high-dose dexamethasone can be used to achieve the same ends. Dexamethasone also reduces the chances of fetal respiratory distress occurring should premature delivery be needed. As a last resort, transsphenoidal surgery during pregnancy can be, and has been, used to decompress the tumor. Since such complications are extremely rare, however, little data have to date been accumulated.