This Article Abstract Full Text (PDF) All Versions of this Article: 27/5/485    most recent Final Manuscript Author Manuscript Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Copyright Permission Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Gillam, M. P. Articles by Colao, A. Search for Related Content PubMed PubMed Citation Articles by Gillam, M. P. Articles by Colao, A.

Advances in the Treatment of Prolactinomas

Division of Endocrinology, Metabolism, and Molecular Medicine (M.P.G., M.E.M.), Northwestern University Feinberg School of Medicine, Chicago, Illinois 60611; and Department of Molecular and Clinical Endocrinology and Oncology (G.L., A.C.), University "Federico II" of Naples, Naples 80131, Italy

Correspondence: Address all correspondence and requests for reprints to: Annamaria Colao, M.D., Ph.D., Department of Molecular and Clinical Endocrinology and Oncology, University "Federico II" of Naples, Via Sergio Pansini 5, Naples I-80131, Italy. E-mail: colao{at}unina.it

	Abstract

Top Abstract I. Introduction II. Observation III. Surgery IV. Radiotherapy V. Medical Therapy VI. Safety of Dopamine... VII. Dopamine Agonist Resistance VIII. Dopamine Agonist... IX. Pregnancy X. Prolactinomas in Children... XI. Special Situations XII. Experimental Therapy XIII. Concluding Remarks References

 
Prolactinomas account for approximately 40% of all pituitary adenomas and are an important cause of hypogonadism and infertility. The ultimate goal of therapy for prolactinomas is restoration or achievement of eugonadism through the normalization of hyperprolactinemia and control of tumor mass. Medical therapy with dopamine agonists is highly effective in the majority of cases and represents the mainstay of therapy. Recent data indicating successful withdrawal of these agents in a subset of patients challenge the previously held concept that medical therapy is a lifelong requirement. Complicated situations, such as those encountered in resistance to dopamine agonists, pregnancy, and giant or malignant prolactinomas, may require multimodal therapy involving surgery, radiotherapy, or both. Progress in elucidating the mechanisms underlying the pathogenesis of prolactinomas may enable future development of novel molecular therapies for treatment-resistant cases. This review provides a critical analysis of the efficacy and safety of the various modes of therapy available for the treatment of patients with prolactinomas with an emphasis on challenging situations, a discussion of the data regarding withdrawal of medical therapy, and a foreshadowing of novel approaches to therapy that may become available in the future.

I. Introduction

A. Historical overview

B. Epidemiology

II. Observation

A. Treatment indications

B. Oral contraceptives for hypogonadism

III. Surgery

A. Surgical approaches

B. Endoscopy

C. Intraoperative imaging

D. Surgical success rates

E. Recurrence and long-term cure

F. Predictors of remission and cure

G. Complications

H. Surgical vs. medical therapy

IV. Radiotherapy

A. Delivery of radiotherapy

B. Analysis of radiotherapy studies

C. Efficacy—conventional radiotherapy

D. Efficacy—single-dose stereotactic radiotherapy

E. Selecting the mode of radiotherapy

F. Complications

G. Conclusions with regard to radiotherapy

V. Medical Therapy

A. Pharmacological profile

B. Mechanisms of action

C. Therapeutic profile of dopamine receptor agonists

D. Serotonin receptor antagonists

E. Conclusions with regard to medical therapy

VI. Safety of Dopamine Agonists

A. Bromocriptine

B. Cabergoline

C. Pergolide

D. Quinagolide

VII. Dopamine Agonist Resistance

A. Definition of dopamine agonist resistance

B. Features and mechanisms

C. Unusual presentations

D. Resistance to prolactin-lowering effects

E. Resistance to mass-reducing effects

F. Treatment approaches

VIII. Dopamine Agonist Withdrawal

A. Overview

B. Bromocriptine withdrawal

C. Cabergoline withdrawal

D. Withdrawal of other dopamine agonists

E. Other conditions associated with remission of hyperprolactinemia

F. Authors’ personal experience

G. Practical implications and approaches

IX. Pregnancy

A. Effect of pregnancy on prolactinoma growth

B. Effects of dopamine agonists on fetal development

C. Recommendations for management of prolactinoma in pregnancy

X. Prolactinomas in Children and Adolescents

A. Epidemiology

B. Presentation

C. Treatment strategy

XI. Special Situations

A. Prolactinomas in males

B. Prolactinomas in multiple endocrine neoplasia

C. Giant prolactinomas

D. Malignant prolactinomas

XII. Experimental Therapy

A. Somatostatin analogs

B. Therapy directed against estrogen/ estrogen receptor

C. Prolactin receptor antagonists

D. Gene therapy

E. Nerve growth factor

F. Molecular therapeutics

XIII. Concluding Remarks

	I. Introduction

Top Abstract I. Introduction II. Observation III. Surgery IV. Radiotherapy V. Medical Therapy VI. Safety of Dopamine... VII. Dopamine Agonist Resistance VIII. Dopamine Agonist... IX. Pregnancy X. Prolactinomas in Children... XI. Special Situations XII. Experimental Therapy XIII. Concluding Remarks References

 
A. Historical overview
IN THE LATE 1920s and early 1930s, a number of groups found that pituitary extracts could induce milk secretion. Riddle et al. (1, 2, 3) found that this substance, which they named prolactin (PRL), was different from other known growth- and gonadal-stimulating substances. They found that this PRL stimulated production of a milk-like substance from the crop sacs of pigeons and doves, and they developed the pigeon crop sac assay (1, 2, 3) that became the standard assay procedure for PRL over the next 30 yr. Eventually, this assay was replaced by specific RIAs in a number of species.

In humans, however, because of the high lactogenic activity of even very highly purified preparations of human GH (4, 5), it was impossible to separate human PRL from GH using the relatively crude pigeon crop assay, and the very existence of a distinct PRL isoform in humans was questioned (5). On the other hand, at the same time, it was observed that most patients with pituitary tumors in whom galactorrhea and amenorrhea were the cardinal clinical features did not have acromegalic features (6) and patients who were known to have isolated, congenital GH deficiency were able to lactate postpartum (7). In 1970, Frantz and Kleinberg (8) developed a sensitive bioassay in which they used excess antibody to GH to neutralize any potential lactogenic effects it had and, for the first time, were able to demonstrate measurable PRL levels in women with puerperal and nonpuerperal galactorrhea but not in most normal men. Subsequently, further purification of human PRL led to the development of RIAs that could finally measure PRL levels in the sera of normal individuals (9).

At the same time that PRL was initially being characterized in the early 1930s by Riddle et al. (1, 2, 3), so too were appearing the first clinical reports of a syndrome of amenorrhea coupled with galactorrhea (10, 11). Over the ensuing 20 yr, three distinct clinical syndromes were described: 1) the Chiari-Frommel Syndrome—amenorrhea, galactorrhea, and low urinary gonadotropins occurring postpartum (12); 2) the Ahumada-Argonz-del Castillo syndrome—nonpuerperal amenorrhea, galactorrhea, and low urinary gonadotropins with no evidence of a pituitary tumor on standard skull x-rays (13); and 3) the Forbes-Henneman-Griswold-Albright syndrome—nonpuerperal amenorrhea, galactorrhea, and low urinary gonadotropins in association with a chromophobe adenoma (6). Overproduction of PRL was postulated in both of the last two syndromes (6, 13). Friesen et al. (14) then demonstrated elevated radioimmunoassayable PRL levels in the serum of a patient with a prolactinoma, the fall in such levels with partial tumor resection, and the production of PRL by the tumor in vitro. The now recognized insensitivity of standard skull x-rays and a better understanding of the pathophysiology of prolactinomas have rendered obsolete this early eponymic classification of PRL disorders.

B. Epidemiology
Prolactinomas are the most frequent pituitary tumors, with an estimated prevalence in the adult population of 100 per million population (15). However, recently Beckers et al. (16) found a much higher prevalence at 55 per 71,000 (775 per million) inhabitants in Belgium. Their frequency varies with age and sex, occurring most frequently in females between 20 and 50 yr old, when the ratio between the sexes is estimated to be 10:1. After the fifth decade of life, the frequency of prolactinomas is similar in both sexes (17, 18). In the pediatric/adolescent age, prolactinomas are rare, but represent about half of all pituitary adenomas, which, overall, account for less than 2% of intracranial tumors (19, 20). One possible explanation for the increased prevalence of prolactinomas in women may be related to the fact that the clinical presentation in women is more evident, usually the classical amenorrhea-galactorrhea syndrome, whereas men may ignore the symptoms of impotence and decreased libido and the diagnosis is often made when signs of compression due to the tumor develop (17). However, studies comparing the clinical and pathological correlates of growth of these tumors in both sexes are lacking, and a more aggressive course of the disease in men has not been ruled out. Delgrange et al. (21) showed a greater growth potential of macroprolactinomas in men than in women as well as a male preponderance of aggressive forms of the disease (i.e., giant, invasive, and malignant prolactinomas).

Prolactinomas cause gonadal and sexual dysfunction related to the hyperprolactinemia and may cause other symptoms related to the tumor expansion. The major objectives of treating patients with prolactinomas are: 1) to suppress excessive hormone secretion and its clinical consequences, such as infertility, sexual dysfunction, and osteoporosis; 2) to control tumor mass, thereby relieving visual field defects, cranial nerve function, and possibly hypopituitarism; 3) to preserve or improve residual pituitary function; and 4) to prevent disease recurrence or progression. Treatment goals for micro- and macroprolactinomas are similar; however, for the majority of microprolactinomas, control of tumor size is a less clinically important endpoint because microprolactinomas do not cause neurological defects, nor are they at substantial risk for growth over time. In larger macroprolactinomas that are at risk for neurological sequelae, control of tumor growth or reduction in size takes priority over the treatment of hypogonadism. Furthermore, patients with macroprolactinomas and hypopituitarism should receive standard hormonal replacement therapy for hypopituitarism as in other types of macroadenomas. The initiation of GH replacement should be delayed if there is a reasonable likelihood that hypopituitarism will resolve and/or normoprolactinemia can be achieved, because GH deficiency can resolve with pharmacotherapy-induced PRL normalization (22). If substantial tumor reduction is not observed after initiating treatment for the prolactinoma and reversal of GH deficiency does not appear realistic, GH therapy should be considered. Because of a theoretical risk of GH/growth factor-induced tumor enlargement, GH therapy in this situation must be done cautiously.

This review provides a critical analysis of the efficacy and safety of the various modes of therapy available for the treatment of patients with prolactinomas. Special attention will focus on challenging situations, including the treatment of patients resistant to dopamine agonists, the treatment of children and adolescents, and management of women who become pregnant. We will review the evidence of the efficacy of withdrawal from dopamine agonists after long-term treatment and, finally, highlight experimental therapies currently under investigation.

	II. Observation

Top Abstract I. Introduction II. Observation III. Surgery IV. Radiotherapy V. Medical Therapy VI. Safety of Dopamine... VII. Dopamine Agonist Resistance VIII. Dopamine Agonist... IX. Pregnancy X. Prolactinomas in Children... XI. Special Situations XII. Experimental Therapy XIII. Concluding Remarks References

 
A. Treatment indications
Asymptomatic patients with prolactinomas do not have an absolute requirement for treatment of their prolactinomas. Indications for therapy in patients with prolactinomas may be divided into two categories: 1) effects of tumor size, and 2) effects of hyperprolactinemia (Table 1). Studies examining the natural history of untreated microprolactinomas have shown that significant growth of these tumors is uncommon (Table 2). Six series of patients with microadenomas who were found to have computed tomography (CT) or tomographic evidence of prolactinomas were observed without treatment for a period up to 8 yr (23, 24, 25, 26, 27, 28). Of 139 women, only nine (7%) microprolactinomas had evidence for growth. Thus, the simple argument that therapy is indicated for a microadenoma to prevent it from growing is fallacious. On the other hand, an untreated prolactinoma should be followed closely to determine whether it is enlarging. It is very unlikely for a prolactinoma to grow significantly without an increase in serum PRL levels, although this phenomenon has been reported (29). Therefore, most patients with microadenomas verified by imaging may be monitored with serial PRL levels. Although there is no consensus on the frequency of imaging after the achievement of normoprolactinemia, most clinicians monitor with pituitary magnetic resonance imaging (MRI) periodically to verify the absence of tumor growth and to ensure that PRL levels are reliable indicators of tumor size. If PRL levels rise or symptoms of mass effects develop (such as headaches), then repeat scanning is indicated to evaluate for the possibility of significant tumor growth. Significant increases in PRL levels usually, although not always (30), reflect tumor growth. A microadenoma with documented evidence of growth demands therapy for the size change alone, because it may be one of the 7% that will grow to be a macroadenoma.

Other indications for therapy are relative, being due to the hyperprolactinemia itself. These include: decreased libido, menstrual dysfunction, galactorrhea, infertility, hirsutism, impotence, and premature osteoporosis. Eugonadal women with nonbothersome galactorrhea do not have specific reasons for therapy. On the other hand, a woman harboring a prolactinoma with amenorrhea and anovulation who wishes to become pregnant has a clear indication for therapy. However, if such a woman does not wish to become pregnant, then therapy may be warranted to prevent osteoporosis or to improve libido.

B. Oral contraceptives for hypogonadism
Hypogondal women with microprolactinomas may be treated for their hypogonadism with oral contraceptives, provided that their PRL levels do not increase substantially and there is no evidence of tumor enlargement (31). Series of patients with prolactinomas who are treated with oral contraceptives for hypogonadism have not shown substantial risk for tumor enlargement (32). Individual case reports of tumor enlargement during estrogen therapy have been documented, but whether tumor enlargement in these cases was related to use of estrogen or reflected the natural progression of these particular tumors is not known. Because of this uncertainty, it is advisable to monitor patients who use oral contraceptives carefully with periodic measurement of PRL levels (33, 34, 35).

The ability to follow patients closely with PRL levels, MRI scans, and bone mineral density (BMD) studies and the knowledge of the efficacy of various modes of therapy allow a highly individualized way of managing patients and choosing the proper mode of their therapy.

	III. Surgery

Top Abstract I. Introduction II. Observation III. Surgery IV. Radiotherapy V. Medical Therapy VI. Safety of Dopamine... VII. Dopamine Agonist Resistance VIII. Dopamine Agonist... IX. Pregnancy X. Prolactinomas in Children... XI. Special Situations XII. Experimental Therapy XIII. Concluding Remarks References

 
A. Surgical approaches
Historically, surgical resection of prolactinomas was the preferred mode of therapy until the mid-1980s, when bromocriptine became available and was shown to be effective in the control of these tumors. Most, although not all, authorities reserve surgical treatment of prolactinomas for special circumstances. A list of surgical indications is presented in Table 3.

When surgery is undertaken, the transsphenoidal approach represents the standard of care for microprolactinomas and the overwhelming majority of macroprolactinomas (40). Craniotomy, which is rarely indicated, is reserved for tumors that are inaccessible via the transsphenoidal approach. Such cases might include patients with large tumors with suprasellar, parasellar, or unusual intracranial extensions, such as those extending toward the frontal or temporal lobes (41). Favorable experience with variations of the standard transsphenoidal approach have recently been described and provide alternatives to transcranial approaches, even for previously unapproachable lesions involving the cavernous sinus and parasellar region (42). Giant and invasive prolactinomas cannot be cured by surgery, regardless of the surgical technique employed or the experience of the neurosurgeon; therefore, if undertaken, the goal of surgery under these circumstances is to debulk with the prospect of improving symptoms related to mass effects (43).

B. Endoscopy
Recent technological advances have catapulted minimally invasive endoscopic techniques to the forefront of transsphenoidal surgery (41). The main advantage afforded by endoscopy is the superior panoramic view. Some surgeons exploit endoscopy as an adjunctive visualizing tool, but increasingly, the endoscope is being used as a stand-alone operating instrument, used in one of several different ways. The most common method consists of an endoscopic endonasal unilateral approach, permitting removal of the entire tumor through the nostril. In addition to the wide surgical view, endoscopy provides three additional potential benefits, including: 1) avoidance of submucosal transseptal dissection, thus eliminating nasoseptal perforations; 2) less patient discomfort due to the lack of nasal packing; and 3) reduced operative time and hospital stay (44). For the neurosurgeon, the major obstacle to adopting the endoscopic technique is that of the need to acquire novel surgical skills (45, 46). Although the number of reported cases and follow-up duration are limited, preliminary studies suggest that complication rates of the endonasal endoscopic approach may be comparable to, or perhaps slightly lower than, those observed using the traditional operating microscope (47, 48, 49, 50, 51). Two studies comparing outcomes of endonasal endoscopic microsurgery and sublabial microsurgery specifically for prolactinomas found no differences in initial surgical remission rates but fewer minor complications associated with endoscopic or endoscopic-assisted procedures (52, 53).

C. Intraoperative imaging
Currently, intraoperative imaging is used by most neurosurgical centers, both for guidance to reach the sphenoid sinus and sella and to assess the extent of tumor resection. C-arm videofluoroscopy is the most widely used guidance device. Frameless stereotaxy, or "neuronavigation", is a newer intraoperative computer-guided imaging system that assists the surgeon in the identification of anatomical landmarks and allows referral to preoperative MRI images in several planes of view simultaneously. This technique is achieved by segmentation and three-dimensional reconstruction of the tumor and adjacent structures. Neuronavigation is most valuable in reoperations where the anatomy may be distorted, but it is also helpful in surgeries involving large, invasive or suprasellar tumors and in cases where the carotid arteries are closely approximated or kinked (40, 54). Because the morphology of the tumor and neurovascular structures shift during tumor resection, frameless stereotaxy cannot be used to estimate the extent of tumor removal. Moreover, these systems have the disadvantage of requiring greater set-up time. Whether the use of neuronavigational systems improves surgical cure rates or reduces the frequency of complications is unknown.

Intraoperative MRI has been used in some centers to delineate tumor borders and to accurately monitor the extent of tumor resection (55). This imaging modality is in an evolving stage of development and improvement. A comprehensive evaluation of the different available MRI systems has been reported by Albayrak et al. (55). The predominant opinion is that intraoperative MRI will have limited use in resecting purely sellar microadenomas but may be beneficial for assessing the adequacy of resection of supra- and parasellar extensions (40). Long-term results are not available to determine whether this application improves surgical outcomes.

Use of ultrasonography is chiefly used in transcranial procedures performed in the resection of giant macroadenomas. The advantage of ultrasound over MRI is the provision of real-time feedback to the surgeon during tumor removal. Because indications to resect giant invasive prolactinomas requiring craniotomy are extraordinarily rare, this imaging modality, in its current form, is suspected to have limited application for the surgical management of prolactinomas.

Regardless of the innovative operative techniques and imaging modalities emerging, the specific surgical treatment plan of a patient with a prolactinoma ultimately depends upon the availability, familiarity, and expertise of these instruments for the neurosurgical center and neurosurgeon to which one refers their patients. Definitive data on the impact of these advances on surgical outcomes are not yet available.

D. Surgical success rates
Surgical outcomes are highly dependent upon the expertise and experience of the neurosurgeon, as well as the size of the tumor. Surgical results from 50 published series are summarized in Table 4. Only results from the latest series from a given neurosurgical/endocrine team are included, omitting data from earlier studies. Criteria for inclusion in this analysis consists of the following: 1) cure or remission rates are reported with respect to size of the tumor (microadenoma vs. macroadenoma); 2) normalization of PRL levels defines surgical remission; and 3) surgical cure rates are reported on the basis of the number of patients with documented follow-up. Combining data from all 50 series, 1596 of 2137 (74.7%) microadenomas and 755 of 2226 (33.9%) macroadenomas were classified as achieving initial surgical remission, i.e., having PRL levels normalized by 1–12 wk after surgery. Within these series, the surgical success rates were highly variable. For series with at least 10 patients, the surgical remission rate varied from 38 to 100% for microadenomas and from 6.7 to 80% for macroadenomas. Similar data were obtained from a mail survey of 80 neurosurgeons, which found surgical cure rates of 74% of 1518 PRL-secreting microadenomas and 30% of 1022 PRL-secreting macroadenomas (56). In this latter report, criteria for the assignment of patients to micro- or macroadenoma status was made on the basis of imaging and/or PRL levels (< or > 200 ng/ml, respectively). Clearly, for the macroadenomas the success rate in large part was dependent on the size of tumors chosen for surgery. In many series, the objective was, appropriately, debulking of a very large tumor rather than cure, and in other series very large tumors were not operated upon.

In a number of series, it was the impression that PRL levels were more predictive of surgical success than actual size of the tumor. Patients with serum PRL levels above 200 ng/ml were found to have a decreased chance for cure at surgery even when stratified within micro- and macroadenoma groups (57, 58, 59, 60, 61, 62). Thus, PRL levels above 200 ng/ml appear to be a risk factor for poor surgical outcome independent of tumor size. An obvious explanation for this finding is lacking, because one would expect higher PRL levels in more highly differentiated tumors, which might thereby impart a greater likelihood of complete tumor resection. It is unknown whether the extent of dural invasion and the degree of histological differentiation of a tumor correlate with PRL levels in prolactinomas.

Gonadal function is almost uniformly restored in both sexes upon achievement of normoprolactinemia after successful surgical resection (63, 64, 65, 66). In young women, normal LH pulsatility is restored as early as the eighth postoperative day (66, 67). Often normal reproductive function is obtained even with PRL levels slightly above normal, but because such patients appear to have a much greater chance of recurrence of more significant hyperprolactinemia (see Section III.F), they cannot be deemed definitively cured. Patients with macroadenomas of all types may be hypopituitary before surgery and, because of the extent of surgery sometimes performed, may have significant changes in pituitary function postoperatively. In an analysis of 84 patients with macroadenomas (36 were prolactinomas), Nelson et al. (68) found that of those with normal preoperative pituitary function, only 78% retained normal function postoperatively. One third with some pituitary deficits before surgery improved, and one third with such deficits had worsened pituitary function after surgery. None of the panhypopituitary patients improved after surgery (68).

E. Recurrence and long-term cure
One of the most controversial areas regarding the surgical management of prolactinomas involves the likelihood of a recurrence of hyperprolactinemia in patients who have undergone an initial remission. Rates of recurrence, as observed with rates of surgical remissions, are highly variable among neurosurgical centers, ranging from 0 (69) to 50% (70). In part, this reflects differences in the level of neurosurgical expertise. Unfortunately, the surgical literature is confounded by variable follow-up times, drop-out rates, and definitions of cure/recurrence. It is possible, and even likely, that surgical series with relatively short follow-up times will underestimate the true recurrence rate because the time to recurrence of hyperprolactinemia in some tumors may be lengthy (71). As for surgical cure, most often, recurrence is defined as the discovery of an elevated PRL level at any point in the postoperative surveillance period after an initial surgical remission. On the other hand, some authors use less stringent criteria, regarding patients with mild asymptomatic hyperprolactinemia as in remission (72). Given all of these factors, a true assessment of recurrence rates is difficult to establish.

Adding further confusion to the controversy are series reporting that recurrence of mild hyperprolactinemia in some women after adenomectomy for prolactinomas resolves with time, and therefore may not definitively reflect operative failure (73, 74). One important study described the course of eight patients who developed recurrent hyperprolactinemia 2 to 10 yr postoperatively and who were monitored thereafter without treatment (74). One patient was found to have primary hypothyroidism and became normoprolactinemic with L-T₄ therapy, thus accounting for her PRL elevation. Of the remaining seven patients, four underwent a spontaneous remission of hyperprolactinemia. This course of relapse, followed by a second remission, has not been widely reported. It is unknown whether these experiences are atypical or whether they reflect a more frequent outcome that has simply gone unrecognized thus far.

From the series compiled in Table 4, recurrence rates for microadenomas [147 of 809 (18.2%)] and macroadenomas [106 of 465 (22.8%)] are similar. It should be emphasized that under most circumstances, the recurrence is detected biochemically (hyperprolactinemia), not necessarily with radiographic documentation of tumor regrowth. Recurrence of the hyperprolactinemia is usually accompanied by sexual/reproductive dysfunction, which thereby serves as an indication for medical therapy to reduce PRL hypersecretion. In a series of patients with microadenomas operated upon by Dr. Jules Hardy, of 58 patients with a normal PRL postoperatively, 25 had a relapse of hyperprolactinemia after a mean of 3.3 yr, but only 10 of these 25 had a recurrence of amenorrhea or galactorrhea, and CT scans showed evidence of a recurrence of the microadenoma in only two (75).

Overall long-term surgical cure rates may be calculated based upon series that have reported both initial remission rates and recurrence rates, understanding that these numbers reflect somewhat of a reporting bias, because they are derived from neurosurgeons who are willing to publish their data. Based on the initial remission rate of 74.7% and the recurrence rate of 18.2% cited above, an overall long-term surgical cure rate for microadenomas among these selected series, using a normal PRL level as the criterion, is 61.1%. For patients with macroadenomas, with an initial remission rate of 33.9% and a recurrence rate of 22.8%, the long-term cure rate is 26.2%. These general numbers may be given to patients when counseling them with respect to choices of therapy. Neurosurgeons who have compiled their own data on surgical cures and remission may provide patients with more meaningful personal estimates for cure, based upon their individual statistics. For patients with giant prolactinomas and those with considerable cavernous sinus invasion, the chance for surgical cure is essentially zero (76, 77, 78).

F. Predictors of remission and cure
A number of studies have analyzed factors that might predict initial surgical remission and likelihood for long-term cure. As discussed above, several studies have identified an inverse relationship between preoperative PRL levels and chances for initial surgical remission (57, 60, 61, 62, 72, 79, 80, 81). Initial surgical success is also correlated with adenoma stage (58, 72). A low immediate postoperative PRL level has been shown to be an excellent predictor of long-term surgical cure (82, 83). For example, one large analysis of surgical outcomes of 339 surgically resected prolactinomas showed that when a postoperative PRL level below 5 ng/ml was achieved, 80.5% of the cohort remained in remission over a mean follow-up of 9.2 yr (83). ¹ In this series, the immediate postoperative PRL level was a better predictor of long-term surgical cure than the preoperative PRL level. A separate (5-yr follow-up) study found that a postoperative PRL below 10 ng/ml predicted biochemical cure with 100% accuracy for both micro- and macroadenomas; cure was unlikely to be obtained in patients with postoperative PRL levels between 10 and 20 ng/ml (i.e., the high normal range) (82). Repeat transsphenoidal surgery for persistent tumor after failed surgery or radiotherapy is curative in less than 50% of the cases (83).

An extensive debate regarding the effects of pretreatment with dopamine agonists on surgical outcomes for prolactinomas has pervaded the surgical literature in the past. Landolt et al. (84) were the first to report better surgical cure rates for microadenomas that had not been exposed to dopamine agonists before surgery. In the hands of some neurosurgeons, it is the impression that dopamine agonist treatment induces tumor fibrosis, creating a tough tumor consistency that made surgical removal difficult. However, the majority of other series investigating this issue have not corroborated these results (72, 85, 86, 87, 88, 89). Because surgery for prolactinomas in the present era most often follows a trial of dopamine agonist therapy, the decision of whether or not to use a dopamine agonist has already been made.

G. Complications
Complications from transsphenoidal surgery for microadenomas are quite infrequent, the mortality rate being at most 0.6%, the major morbidity rate being about 3.4% (visual loss 0.1%, stroke/vascular injury 0.2%, meningitis/abscess 0.1% and oculomotor palsy 0.1%) and cerebrospinal fluid (CSF) rhinorrhea occurring in 1.9% (41, 56, 90, 91). The mortality rate for transsphenoidal surgery for all types of secreting and nonsecreting macroadenomas is 0.9%, the major morbidity rate is 6.5% (visual loss 1.5%, stroke/vascular injury 0.6%, meningitis/abscess 0.5%, and oculomotor palsy 0.6%), and the rate of CSF rhinorrhea is 3.3% (41, 56, 90, 91). Transient diabetes insipidus is quite common with transsphenoidal surgery for both micro- and macroadenomas, and permanent diabetes insipidus occurs in about 1% of surgeries on macroadenomas (56, 91). Hypopituitarism is common in patients with macroadenomas before surgery as a result of mass effects, occurring in more than 50% of patients. With surgery, either further worsening or improvement may occur (68). GH deficiency after transsphenoidal surgery for microprolactinomas was reported as high as 30% in one series (92). Surgery involving craniotomy is much more hazardous. Although visual field defects and reduction in visual acuity can be improved in 74% of patients whose macroadenomas abut the optic chiasm (93), a small number of patients with normal visual fields may have a reduction of vision after surgery due to herniation of the chiasm into an empty sella, direct injury or devascularization of the optic apparatus, fracture of the orbit, postoperative hematoma, or cerebral vasospasm (94).

H. Surgical vs. medical therapy
Although most authorities recommend transsphenoidal surgery as a second line option for prolactinomas, some experts continue to advocate surgery as a potentially curative procedure in selected patients (72, 92, 95, 96). These experts propose that young patients with microadenomas who have a good chance of cure could avoid the need for extended medical therapy. Enthusiasm for this perspective is tempered by recent evidence that withdrawal of cabergoline may lead to sustained remission in some cases (97, 98). Couldwell and Weiss (99) make an economic argument that surgical intervention is less expensive than lifelong medical therapy. On the other hand, a surgery complicated by hypopituitarism has the potential to incur significantly higher costs (with replacement of pituitary hormones) and lead to higher morbidity (secondary adrenal insufficiency) over many years. Admittedly, hypopituitarism is an unlikely complication for the resection of a microprolactinoma in the hands of an experienced neurosurgeon. One should also bear in mind that with the expected upcoming availability of cabergoline in generic form, even an uncomplicated transsphenoidal surgery may prove to be more expensive than the cost of medical therapy. Given its efficacy and safety, as discussed in Section V, medical therapy remains the first line therapy for the treatment of prolactinomas.

	IV. Radiotherapy

Top Abstract I. Introduction II. Observation III. Surgery IV. Radiotherapy V. Medical Therapy VI. Safety of Dopamine... VII. Dopamine Agonist Resistance VIII. Dopamine Agonist... IX. Pregnancy X. Prolactinomas in Children... XI. Special Situations XII. Experimental Therapy XIII. Concluding Remarks References

 
A. Delivery of radiotherapy
The availability of highly effective medical and surgical therapies for the majority of prolactinomas has rendered the role of radiotherapy in the management of prolactinomas as one of adjunctive therapy. In most cases, radiotherapy is used after failed transsphenoidal surgery and medical therapy. Rarely, in a few centers, it has been administered postoperatively as a prophylactic measure to prevent growth of a remnant tumor.

Today, several methodologies for the delivery of radiotherapy are available. Conventional fractionated external beam radiotherapy involves the use of several ports to concentrate an x-ray beam on the pituitary fossa by a crossfire technique while the patient is immobilized in an individually shaped plastic mask. Supravoltage radiotherapy is delivered in daily doses of 200 cGy 4–5 d/wk over a period of 5–6 wk up to a total dose of 4500–5000 Gy (100). Stereotactic conformal radiotherapy (SCRT) is also a fractionated form of radiotherapy, but uses stereotactic techniques to deliver radiation with higher precision. The underlying principle of SCRT is to shape the radiation beams to conform to the shape of the tumor, thereby reducing radiation exposure to surrounding normal brain (101). Most SCRT is delivered with linear accelerators (LINAC) that generate photon beams focusing on a stationary target, using a moving gantry system (102). Most recently, single dose radiotherapy has become widely available and is being increasingly used. This form of radiotherapy delivers a necrotizing dose to the tumor, which has been stereotactically defined using three-dimensional image processing. The hallmark of this type of radiotherapy is the sharp dose gradient of radiation at the treatment field edges, which reduces the dose of radiation to the surrounding normal brain tissue. Most single-dose radiotherapy uses cobalt-60 gamma radiation emitting sources ("gamma-knife" radiotherapy) arranged in a hemisphere to focus on a central target. This results in multiple small radiation spheres that are combined in a multiple isocenter technique to conform to the shape of nonspherical pituitary tumors (103). An alternative method of delivering single-dose radiation therapy is with the use of a LINAC-based system that has been modified to limit mechanical instability and inaccuracy (104, 105). There is no clear advantage for either of these single-dose techniques (gamma knife vs., LINAC) in terms of their ability to spare normal tissue from high radiation doses (106). Treatment of pituitary adenomas with single-dose radiation therapy using heavy-charge particle proton beams is very limited, because few centers have the facilities to provide this form of radiotherapy, due to its high operational and maintenance costs (107). As of yet, there are no adequate studies to conclude whether there is one mode of single dose radiotherapy that has superior efficacy or safety.

Regardless of the delivery system, the aim of all high precision techniques is to minimize radiation exposure to the surrounding normal tissue. The combination of better immobilization and high-definition three-dimensional imaging has been the most important determinant of improvements in modern radiotherapy—less so the technique of delivery (103). The two general therapeutic goals for performing radiotherapy in prolactinomas are: 1) arrest of tumor growth and prevention of further problems from mass effects; and 2) normalization of hyperprolactinemia (108, 109, 110, 111, 112). However, it is possible that persistent postradiotherapy elevations of PRL in patients with prolactinomas are due to a radiation effect and are not caused by hypersecretion from residual tumor (113). Although these arguments may hold true, the variable definitions of "mildly elevated PRL levels" and "improvement" preclude an accurate assessment of radiotherapy efficacy.

B. Analysis of radiotherapy studies
There are several important caveats regarding the critical interpretation of data in clinical studies of the efficacy of radiotherapy for the treatment of prolactinomas. The first relates to the definitions of tumor control endpoints. Tumor control may be defined with either endocrinological (normalization of PRL levels) or volumetric (long-term radiographic assessment of tumor size) parameters. Many series report a tumor "control" rate, which often refers to the stable PRL levels or the absence of radiographic progression. Moreover, many of the series do not report the mean or median duration of follow-up. Under these circumstances, the data do not provide an appropriate measure of efficacy. The second important observation is that the majority of the studies are retrospective, single-arm analyses. Few of them report the rate of decline of PRL. Finally, and most importantly, many of the studies are confounded by the inclusion of patients who were receiving concomitant medical therapy. A number of patients in whom cure was reported continued to require dopamine agonist therapy. Therefore, in some cases, it is impossible to separate out any PRL-lowering effects of radiotherapy from the effects of dopamine agonists. In the present review, we have chosen to confine the analysis to the normalization of hyperprolactinemia. It should nevertheless be acknowledged that in many cases the goal of radiotherapy for treatment-resistant tumors was control of growth or alleviation of mass effects. To accurately assess the efficacy of radiotherapy for these situations, standard criteria for growth control would be necessary.

C. Efficacy—conventional radiotherapy
Approximately 250 patients have been reported who have undergone treatment with conventional radiotherapy alone, or after failure of medical and/or surgical therapy (114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126). In all of these settings, normalization of hyperprolactinemia was infrequent, with an overall normalization rate for the entire series of 34.1%. When normal PRL levels were achieved, it was only with an extended latency in most cases.

Conventional radiotherapy after noncurative surgery rarely normalizes PRL levels. A total of three studies report PRL normalization rates of 1 of 11 (9%) (119), 0 of 11 (0%) (115), and 3 of 13 (23%) (121) when radiotherapy was administered after incomplete surgical resection. Two series of patients who received LINAC-based fractionated radiotherapy after unsuccessful transsphenoidal surgery achieved PRL normalization rates similar to those receiving conventional fractionated radiotherapy (36.3 and 25%) (127, 128).

The remainder of the studies in this group included patients treated with a dopamine agonist or with all three modalities (surgery, dopamine agonists, and radiotherapy). Thus, it is uncertain what specific effect radiotherapy had in PRL lowering, or normalization, in these patients. Because these prolactinomas often represent the most therapy-resistant tumors, normalization of hyperprolactinemia may not have been feasible. Therefore, the rates of normalization are expected to be poor. In many of these cases, the goal of radiotherapy may have been to control further growth or to relieve mass effects on cranial nerves.

D. Efficacy—single-dose stereotactic radiotherapy
Almost 300 patients (Table 5) have been reported who have undergone treatment with single-dose stereotactic radiotherapy alone, or after failure of medical and/or surgical therapy (127, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149). In all of these settings, normalization of hyperprolactinemia was infrequent, with an overall normalization rate for the entire series of 31.4%. One could argue that the short follow-up duration among these series may have underestimated the complete response rate for these tumors treated with single-dose radiotherapy.

A second notable study by Landolt and Lomax (138) reported the outcomes of 20 patients who underwent gamma-knife radiosurgery after unsuccessful transsphenoidal surgery and/or "failed" medical therapy. Normoprolactinemia was achieved in five patients, all of whom were able to discontinue medical therapy. For 11 patients, PRL levels normalized or declined by at least 20%, but only with continuation of medical therapy. These subjects were regarded as "improved", although clearly in the absence of a control group, any improvement attributed to the effects of radiotherapy cannot be distinguished from effects of medical therapy. Furthermore, the clinical significance of a 20% decline in hyperprolactinemia is uncertain, because it would not be expected to alter dopamine agonist therapy continuation or dosage under these circumstances. Treatment with radiosurgery failed entirely in four patients. Therefore, a 25% complete response rate was achieved for gamma-knife radiotherapy in this series.

E. Selecting the mode of radiotherapy
The advantages and disadvantages of various modes of radiotherapy are important considerations when referring a patient with a prolactinoma to a radiotherapist. Although it is true that a large single dose of radiation is more effective in cell death than the same dose delivered in several smaller fractions, large single doses of radiation are more toxic to normal tissue than similar doses given in a fractionated manner. In the early days of single-dose radiotherapy, high doses for large pituitary adenomas near the optic apparatus resulted in a high incidence of optic neuropathy (150). The risk of damage is dose dependent, with a 78% risk of optic neuropathy in patients receiving more than 15 Gy and a 27% risk for those receiving 10–15 Gy to the optic apparatus (151, 152). To achieve an acceptable fall-off gradient with single-session therapy, current practice aims at limiting irradiation of the optic apparatus to single doses of less than 8 Gy (103, 112, 113, 139). As a result, pituitary adenomas with significant suprasellar extension, or those with less than 5-mm clearance between the tumor margin and the optic apparatus are poor candidates for single-dose radiotherapy (103, 112, 139). On the other hand, tumors with cavernous sinus invasion can be good candidates for single-dose radiotherapy, because the cranial nerves in the cavernous sinus are relatively radio-resistant (112, 153, 154). Fractionated radiotherapy is also preferable to single-dose radiotherapy when the tumor volume is so large (>3 cm) that an effective radiation dose cannot be safely delivered in a single session (112, 113, 139). It is estimated, based upon extrapolation from an integrated logistical formula, that the maximum diameter of a spherical tumor that is treatable with a risk of complications less than 3% is approximately 35 mm (155).

Prospective evaluations of the rate of PRL normalization in head-to-head comparisons of fractionated vs. single-dose radiotherapy for prolactinomas have not been published. Even if such series were to exist, one would need to critically analyze the comparability of the tumors in each group. For conventional radiotherapy, there is no restriction as to the size of the tumor or the proximity to the optic apparatus. As noted above, prolactinomas treated with single-dose radiotherapy will largely consist of intrasellar adenomas well away from the optic apparatus. In contrast, prolactinomas treated with fractionated radiotherapy consist of not only intrasellar adenomas, but also large adenomas with suprasellar extension that may lie in close proximity to the optic apparatus. Complete hormonal and tumor growth responses are undoubtedly more difficult to achieve in this latter category of tumors.

In addition to patient convenience, one of the proposed advantages of single-dose radiotherapy over fractionated radiotherapy is its shorter latency to hormonal and tumor size responses. It appears likely that single-dose radiotherapy lowers PRL levels more rapidly than conventional external beam radiotherapy, although prospective comparative studies are not available to verify this claim. Six of the 21 series listed in Table 5 reported the latency to normalization of PRL levels for their patients with prolactinomas treated with single-dose radiotherapy; these responses ranged from 1 to 2 yr (129, 137, 139, 147, 148, 149). The latency to normalization of hyperprolactinemia for conventional external beam radiotherapy is on the order of several years.

F. Complications
The most frequent long-term morbidity of conventional radiotherapy is radiation-induced hypopituitarism, with a cumulative actuarial risk of approximately 50% at 10–20 yr (111, 156, 157). Hypopituitarism is likely secondary to hypothalamic and pituitary damage, although the former is considered of primary importance (158). Recently, it has been discovered that the consequences of hypopituitarism may be more significant than issues related to hormone replacement dosing and monitoring. A recent large prospective study from the United Kingdom showed that the standardized mortality rate was higher in patients with hypopituitarism that had been treated with radiotherapy compared with those who had not received radiotherapy (159). A large proportion of this excess was due to a significant increase in cerebrovascular disease-associated deaths in the radiotherapy group.

Additional complications that occur months to years after radiotherapy of pituitary adenomas include cerebrovascular accidents, optic nerve damage, neurological dysfunction, and soft tissue reactions (157, 160, 161, 162). Conventional radiotherapy is associated with an increased risk of secondary radiation-induced intracranial malignancies, with a cumulative risk of 2.0% at 10 yr and 2.4% risk at 20 yr (163, 164, 165).

The incidence of hypopituitarism after single-dose stereotactic radiotherapy is difficult to establish at present. The reported rates of hypopituitarism vary widely, ranging from 0–36%, after single-dose radiotherapy (132, 134, 136, 141, 142, 143, 148). These analyses are confounded by factors such as previous pituitary surgery in some individuals. A long-term follow-up study with a mean follow-up of 17 yr determined a relatively high cumulative incidence of hypopituitarism at 72% (108). Cranial neuropathies have been reported after single-dose radiotherapy. A high incidence of optic neuropathies was reported in the early use of single-dose therapy when higher doses were administered. With dose changes and technical improvements, these adverse effects are less frequent. The risk of damage to the optic apparatus is approximately 1%. The severity of these cases ranges from nonspecific visual loss to blindness (113, 148, 166). Cranial neuropathies involving nerves that traverse the cavernous sinus (III, IV, V, VI) are less common and often transient (113, 153, 166). Radiation necrosis of surrounding brain tissue occurs in approximately 0.2–0.8% of cases (113, 160). As of yet with limited follow-up, there have been no reported cases of secondary intracranial malignancies who have undergone single-dose radiotherapy.

G. Conclusions with regard to radiotherapy
Overall, radiotherapy has a very limited role in the treatment of patients with prolactinomas. The very high rate of efficacy of dopamine agonists (see Section V) along with the high complication rates of radiotherapy render treatment with this modality rarely necessary. The few patients who now require radiotherapy are those who do not respond to dopamine agonists and who cannot then be cured by surgery. When a large tumor remains after surgery, then conventional radiotherapy is the best modality. However, a small residual tumor, especially when involving the cavernous sinus, may be better treated with stereotactic radiotherapy.

	V. Medical Therapy

Top Abstract I. Introduction II. Observation III. Surgery IV. Radiotherapy V. Medical Therapy VI. Safety of Dopamine... VII. Dopamine Agonist Resistance VIII. Dopamine Agonist... IX. Pregnancy X. Prolactinomas in Children... XI. Special Situations XII. Experimental Therapy XIII. Concluding Remarks References

 
The compounds used in clinical practice to treat prolactinomas are all dopamine receptor agonists. Among these, bromocriptine, cabergoline, pergolide, and quinagolide are the most commonly used. The dopamine agonists lisuride and terguride are less frequently used, as is metergoline, a serotonin antagonist.

A. Pharmacological profile
Bromocriptine, pergolide, and cabergoline are all ergot derivatives. The only nonergot derivative that is used in clinical practice is quinagolide. The chemical structures of the most used compounds are shown in Fig. 1. The ergot-derivative dopamine agonists comprise a group of indole alkaloids that are predominantly found in various species of the ascomycete Claviceps (167). The ergot alkaloids can all be considered derivatives of the tetracyclic ergoline skeleton and can be divided into two main groups based on their structural characteristics (167). The first group includes all lysergic acid derivatives of the acid amide types, such as amine alkaloids (ergonovine) and the structurally more complex ergopeptines (ergotamine, ergocristine). The second group includes the so-called "clavine alkaloid" derivatives that contain either a methyl or a hydroxymethyl group at position 8 (167). The ergot alkaloids and their derivatives have a wide spectrum of pharmacological actions that include central, neurohumoral, and peripheral effects, mediated by norepinephrine, serotonin, and dopamine receptors. The diversity of biological properties of ergot derivatives is likely due to diverse mechanisms of action at the cellular and molecular levels (167). Because ergot derivatives interact with different receptor sites, it is not surprising that the drugs developed (as well as the natural alkaloids) display a number of side effects (167). The clinical application of ergot derivatives in the treatment of other clinical conditions, such as postpartum hemorrhage, migraine, and other vascular headaches, orthostatic hypotension, and Parkinson’s disease will not be discussed here.

View larger version (22K): [in this window] [in a new window]   FIG. 1. Biochemical structure of dopamine agonists. Bromocriptine, pergolide, and cabergoline are all ergot derivatives. Quinagolide is a nonergot derivative.

B. Mechanisms of action
In contrast to the other pituitary hormones, PRL secretion is mainly regulated by the inhibitory tone exerted by dopamine with minor additional inhibitory activity played by -aminobutyric acid and cholinergic pathways (175). TSH-releasing hormone, serotonin, estrogens, endogenous opiates, and vasoactive intestinal polypeptide stimulate PRL secretion, but their role is clearly minor compared with that of dopamine (175).

Classically, dopamine receptors have been divided into D₁ receptors, which stimulate adenylyl cyclase activity, and D₂ receptors, which inhibit this enzyme (176, 177, 178); three further discrete receptor subtypes have been described (D₃, D₄, and D₅) with less activity on PRL secretion (177). Dopamine inhibition of PRL secretion is mediated by the D₂ dopamine receptors expressed by normal and tumorous lactotrophs (176, 177, 178). D₂ receptors belong to the family of G protein-coupled receptors, characterized by a single polypeptide chain containing seven hydrophobic transmembrane domains: besides their effect on adenylyl cyclase, they are able to inhibit inositol phosphate production (179) with an effect that involves G proteins sensitive to pertussis toxin (176, 177, 178, 179). Additionally, dopamine inhibits arachidonic acid release from pituitary cells independently from the other transduction mechanisms (180). Two isoforms generated by alternative splicing of the D₂ dopamine receptor have been described (181). These two isoforms differ by a 29-amino acid additive sequence located within the third intracytoplasmic loop that interacts with G proteins: dopamine inhibition of adenylyl cyclase activity is observed with both isoforms. Stimulation of D₂ receptors by dopamine reduces adenylyl cyclase activity that consequently reduces intracellular cAMP levels in normal as well as in tumoral lactotrophs (182). The inhibition of cAMP levels is a key step in the inhibition of PRL release by dopamine (183). It is likely that all dopaminergic ergot derivatives share similar mechanisms of action (175).

Dopamine agonists reduce the size of prolactinomas by inducing a reduction in cell volume (via an early inhibition of secretory mechanism, and a late inhibition of gene transcription and PRL synthesis), as well as causing perivascular fibrosis and partial cell necrosis (184). There may also be a true antimitotic effect of these drugs. Histologically there is a reduction in secretory activity and cell size, an increase in immunoreactive PRL cellular content and inhibition of exocytosis (185).

C. Therapeutic profile of dopamine receptor agonists
1. Bromocriptine.
More than 25 yr ago, bromocriptine was introduced into clinical practice as the first medical treatment for prolactinomas (186, 187, 188). Bromocriptine-mesylate is a semisynthetic ergot derivative that has D₂ receptor agonist and D₁ antagonist properties. It has a relatively short elimination half-life, so that it is usually taken two or three times daily, although once daily may be effective in some patients. Generally, the therapeutic doses are in the range of 2.5–15 mg/d, and most patients are successfully treated with 7.5 mg or less. However, doses as high as 20–30 mg/d may be necessary for patients who demonstrate resistance. For microprolactinomas bromocriptine is successful in 80 to 90% of patients in normalizing serum PRL levels, restoring gonadal function, and shrinking tumor mass (175). For macroprolactinomas, normalization of serum PRL levels and tumor mass shrinkage occur in about 70% of patients treated with bromocriptine even when given at low doses; visual field defects improve in the majority of patients (175). In most patients, headache and visual field defects improve dramatically within days after the first administration of bromocriptine, with gonadal and sexual function improving even before complete normalization of serum PRL levels. Although prolactinomas usually remain sensitive to bromocriptine, this drug usually does not "cure" these pituitary adenomas, and the withdrawal of therapy often results in recurrent hyperprolactinemia; tumor regrowth may occur later, with the consequent risk for compromised vision (see Section VIII). Prolonged bromocriptine treatment has been associated with increased fibrosis of prolactinomas (189) and with increasing tumor consistency. PRL normalization with bromocriptine is also associated with an increase in bone density both in women (190) and in men (191) and with improvement of semen quality in men (192). Other formulations of bromocriptine, long-acting and the long-acting repeatable forms for im injections, an intranasal powder, and an intravaginal tablet, were developed to overcome side effects such as nausea, vomiting, postural hypotension, and headache (see Section VI). These reactions were considered to be due to the rapid absorption of bromocriptine, which is administered two or three times a day, thus causing high blood levels. However, despite promising data (190, 191, 192, 193, 194, 195), none of these formulations were ever introduced in the pharmaceutical market for hyperprolactinemia. Bromocriptine (as a first-generation dopamine receptor agonist) has been largely superseded by more potent compounds with longer lasting effects and improved side effect profiles. Nevertheless, bromocriptine is still widely used to treat prolactinomas, primarily in young women desiring pregnancy (see Section IX).

2. Cabergoline.
Cabergoline is a D₂ selective agonist widely used to treat prolactinomas. It strongly suppresses PRL secretion both in vivo and in vitro and preliminary studies showed a significant PRL inhibition within 12 h after treatment with cabergoline and bromocriptine in cultured pituitary cells from estradiol-induced rat pituitary tumors (196, 197). Inhibition of de novo PRL synthesis was more pronounced with cabergoline than bromocriptine treatment (196). The continued oral administration of cabergoline significantly reduced both PRL levels and the weight of the pituitary during 15–60 d of treatment as compared with bromocriptine (197). One single dose of cabergoline (0.2–0.6 mg) in healthy male volunteers induced a dose-dependent PRL inhibition (198). In healthy men, single doses of 0.5, 1, and 1.5 mg of cabergoline completely suppressed PRL levels (<1 µg/liter) (199). In healthy women with regular menses, cabergoline at doses of 0.4–0.6 mg induced a 43–76% PRL suppression; PRL levels returned to baseline within 24 h after the low 0.4-mg dose but remained suppressed until 5 d after the administration of 0.6 mg (200).

The beneficial effects of cabergoline in resolving hyperprolactinemia are widely known (201) (Table 6). Significant decreases in serum PRL levels occur in as many as 95% of hyperprolactinemic women during chronic cabergoline treatment at a dose of 1 mg twice weekly (202). In a multicenter, randomized, 24-wk trial conducted in 459 hyperprolactinemic women (203), cabergoline induced normal PRL levels in 83% compared with 59% with bromocriptine; ovulatory cycles or pregnancies were recorded in 72% vs. 52%, and side effects were less frequent, less severe, and shorter lived. In a retrospective study of 455 patients (204), cabergoline treatment normalized PRL levels in 86% of 425 patients with available follow-up (92% of 244 patients with idiopathic hyperprolactinemia or microprolactinoma, and 77% of 181 patients with macroprolactinoma); 13% had side effects but only 4% discontinued cabergoline therapy because of side effects (204). Generally, the median dose of cabergoline at the start of therapy was 1 mg/wk in patients with macroprolactinomas and 0.5 mg/wk in those with idiopathic hyperprolactinemia or microprolactinomas (0.5 mg/wk) (201). A remarkable tumor-shrinking effect of cabergoline has been observed in patients with macroprolactinomas (205): 12–24 months of treatment with cabergoline induced a greater than 20% decrease of baseline tumor size in more than 80% of cases, with complete disappearance of tumor mass in 26–36% of cases. Moreover, Colao et al. (206) showed that cabergoline treatment induced further tumor shrinkage in 60% of patients previously treated with other dopamine-agonists compared with 82.3% of previously untreated patients (Fig. 2A). Cabergoline treatment is also effective and safe in patients with prolactinomas with onset in childhood or adolescence (see Section X). The superiority of cabergoline over bromocriptine was supported by a comparative retrospective study by Di Sarno et al. (207). Based on these data, cabergoline treatment is clearly indicated as the primary approach to macroprolactinomas. Lastly, cabergoline seems to induce fewer side effects than other dopamine agonists (see Section VI). At present, cabergoline is certainly the most effective compound to treatment prolactinomas, with very good patient compliance with long-term treatment regimens.

View larger version (22K): [in this window] [in a new window]   FIG. 2. Tumor shrinkage with cabergoline. A, Cabergoline induced further tumor shrinkage in 60% of patients previously treated with other dopamine-agonists, compared with 82.3% of previously untreated patients. •, De novo patients; , patients intolerant to bromocriptine; , patients resistant to bromocriptine; , patients responsive to a previous treatment course of bromocriptine and treated with cabergoline. Data are derived from Ref. 206 . B, Comparison of tumor reduction responses in micro- and macroprolactinoma with bromocriptine or cabergoline. Data are derived from Ref. 207 .