The Case for Initial Surgical Removal of Certain Prolactinomas

WHEN I was asked to write this overview and introduction to the “Controversy” regarding whether the initial management of prolactinomas should be medical or surgical, I was rather surprised, thinking there really was no controversy. However, Drs. Wilson and Thorner, in the following discussions, both make cogent arguments for surgical and medical management. Fortunately for both endocrinologists and patients, both approaches, in fact, are excellent with high success/morbidity ratios. Prolactinomas are the most common of the hormone secreting pituitary tumors and usually present because of signs and symptoms caused by hyperprolactinemia, i.e. galactorrhea and oligo/amenorrhea in women, impotence in men, and loss of libido and infertility in both sexes. The great majority of prolactinomas are microadenomas, and we have done such a good job of teaching about this disorder that most such patients remain in the hands of the obstetrician/gynecologist or the primary care generalist and are treated medically. Macroadenomas may, in addition, present with mass effects, such as headaches, visual field defects, and hypopituitarism. Most patients with macroadenomas are treated by endocrinologists, we hope. As you will read, Dr. Wilson reports truly superb results from transsphenoidal surgery for PRL-secreting microadenomas. His cure rate of about 90%, with a recurrence rate of 16%, yields a long-term cure rate of 74%. However, not all of us are fortunate enough to have Dr. Wilson as our neurosurgeon. For reasons not entirely clear to me, there appears to be not more than a handful of neurosurgeons in the U.S. who are capable of achieving such results. An argument could be made for referring all of our patients to this handful, but that is not always practical. The clerk in the HMO office says that the HMO has a perfectly good neurosurgeon who does this type of surgery and the patient must be operated upon by the surgeon in his or her HMO. Alternatively, the patient may not wish to fly to a distant city for surgery. A few years ago, I summarized most of the written surgical literature to 1993, finding that 872 of 1224 (71%) patients with PRL-secreting microadenomas appeared to be cured initially by surgery, with a subsequent recurrence rate of 17%, giving a long-term cure rate of only 59% (1). As this is published data, it may well be that the neurosurgeon who does not operate on a sufficient number that would comprise a publishable series may have even less success. Under those circumstances, even a 50% long-term cure rate may be optimistic. For microadenomas the mortality was 0.27% and the morbidity was 0.4% (2, 3), numbers clearly different from zero. Dr. Wilson agrees that for larger macroadenomas the chance for surgical cure is quite low; for men he had an initial cure rate of 38% with a recurrence rate of 40%, giving a long-term cure rate of only 23%. In the review I mentioned above (1), only 400 of 1256 (32%) patients with macroadenomas appeared to be cured initially, with a 19% recurrence rate, giving a long-term cure rate of only 26%. In contrast to these surgical results, Dr. Thorner states that medical therapy with bromocriptine is successful in at least 75% of cases in restoring gonadal function but states that such treatment needs to be continued for the rest of the life of the patient. However, it is clear that in most patients the dose can be reduced over the years (4) and in 10–20% of patients the drug may be stopped altogether without recurrence of hyperprolactinemia (4–7). Recent evidence suggests that cabergoline, a once weekly dopamine agonist, may be even more efficacious and better tolerated for women with microadenomas (8). All four dopamine agonists studied extensively—bromocriptine, pergolide, quinagolide, and cabergoline—appear to be remarkably similar in their abilities to shrink large tumors, with about 70% of such macroadenomas shrinking by 25% or more, an amount sufficient to improve visual fields (9). There is yet another medical therapy, and that is simply using estrogen replacement therapy in women not desiring fertility. Corenblum and Donovan (10) have shown in a small series that such treatment did not increase tumor size in a number of women with microadenomas and in only one with a macroadenoma. I am sure that there will be some tumors that will increase in size when exposed to estrogen, so that women treated in this manner must be followed carefully, especially those with macroadenomas. It should come as no surprise, therefore, that I lean towards Dr. Thorner’s point of view. Perhaps if this article were written in a neurosurgery journal with the introduction by a neurosurgeon, a different perspective might be given. Regardless, it is important for all of us as clinicians to present the data objectively to our patients, citing both advantages and disadvantages of each approach. When I have done this, I have had some patients with microadenomas wish to take a 50/50 chance on long-term cure with surgery and other patients run, not walk, in the opposite direction. It is indeed fortunate that we have two such excellent therapies to choose between. PITUITARY tumors are almost always benign and usually are very slow growing. Patients with pituitary tumors present in one of three ways; either they present with the symptoms of their endocrine dysfunction, with the symptoms of the mass effect of the tumor, or with the diagnosis being made incidentally when an magnetic resonance imaging (MRI), computed tomography scan, or skull × ray was performed for some other reason such as head injury or sinus problems. Once the diagnosis of a pituitary tumor has been made it is important to tailor the therapy to the clinical problem and the needs of the patients. This discussion is focused on prolactinomas; thus it is important to be able to clearly distinguish a prolactinoma from other pituitary tumors. This is important not only for academic purposes but also for practical therapeutic reasons. The initial evaluation to distinguish prolactinomas from other pituitary tumors should include measurement of basal serum prolactin level, thyroxine, thyroid hormone uptake and thyrotropin, and serum insulin-like growth factor (IGF)-1, MRI of the pituitary and, if there is a compression of the optic chiasm, computerized objective evaluation of the visual fields. These tests make it possible to determine the absolute level of the serum prolactin, exclude hypothyroidism as a cause of hyperprolactinemia, and exclude occult or early acromegaly, which may present with many similar signs and symptoms of prolactinomas It is established that there is a relationship between the size of the prolactinoma and the level of serum prolactin. The presentation of hyperprolactinemia differs between men and women. It is more frequently diagnosed in women, although in post mortem studies of the pituitaries of patients who died from nonendocrine disease the prevalence of prolactinomas is similar between the sexes. Women present with amenorrhea or any menstrual abnormality, from oligomenorrhea to menorrhagia, as well as irregularities of the cycle length. They may present with infertility without any obvious abnormality of menstrual function. Galactorrhea occurs in 80% of women with hyperprolactinemia; however of women with galactorrhea, only 50% have hyperprolactinemia. Thus galactorrhea is a poor marker of the presence or absence of hyperprolactinemia. Additional symptoms include dyspareunia and reduced libido caused by hypogonadism. Although post-oral contraceptive amenorrhea is often associated with hyperprolactinemia, multiple studies have demonstrated that taking the oral contraceptives does not increase the incidence of hyperprolactinemia. Men with hyperprolactinemia most commonly present with the mass effect of the tumor-visual field abnormalities, headaches, or external ophthalmoplegia. Alternatively, they present with symptoms of gonadal dysfunction - impotence, oligospermia, or low testosterone level picked up at the time of routine medical examination. The majority of men with hyperprolactinemia are unaware of their sexual dysfunction and rationalize it as an effect of “normal aging.” Their sexual partners are always aware that there is a problem. This may be accounted for by a central nervous system effect of prolactin reducing libido. Galactorrhea is unusual in men and is present only in about 30% of hyperprolactinemic men. However the presence of galactorrhea in men is virtually pathognomonic of hyperprolactinemia. Hyperprolactinemia is unusual before puberty, but does occur. At that stage in development, prolactin levels are often relatively low for the size of the tumor, and the child presents either with delayed puberty, arrest of growth, or symptoms of the mass effect of the pituitary tumor. The objectives of therapy for prolactinomas are: restoration of prolactin levels to normal with return of the eugonadal state; reduction in the size of any large tumor, particularly if it is compressing the optic chiasm or producing other symptoms; restoration of other pituitary functions; prevention of recurrence of the condition. Medical therapy, using dopamine agonist drugs such as bromocriptine, is effective in at least 75% of cases in achieving all these objectives. Thus medical therapy is the first line of treatment for this condition. Transsphenoidal pituitary microsurgery is effective in removing microprolactinomas (diameter < 10 mm) and where the preoperative prolactin level is less than 250 ug/L in a similar percentage of cases. However this depends largely on the experience and skill of the neurosurgeon. Such patients should only be operated on in major centers, where there is a good track record of successful surgery. Surgery is also used in patients who are intolerant of or resistant to medical therapy. Radiation therapy is reserved for patients who have failed medical and surgical therapy. One of the most common problems with medical therapy is intolerance to dopamine agonist therapy, with symptoms of nausea, vomiting, and postural hypotension. Milder side effects include nasal stuffiness and constipation. The physician must explain to the patient that side effects can be avoided by gradual escalation of the dose and that tolerance develops for the side effects but not for the prolactin lowering effect. The medication should always be taken with food since this delays its absorption and reduces or obviates side effects. Adverse effects can usually be avoided by attention to detail in the initiation of such treatment. Bromocriptine is the only dopamine agonist that has been approved by the FDA for the treatment of hyperprolactinemia and infertility. In the U.S.A. pergolide is also available but is not approved for this indication. In other parts of the world there are several other dopamine agonists, including for example quinagolide (CV 205–502) and cabergoline. The longest experience is with bromocriptine, and the others may have some marginal advantages, including possibly few initiating side effects with quinagolide, longer duration of action of pergolide requiring only once daily administration, and even longer duration of action of cabergoline requiring only once or twice weekly administration. In this country, for women in the reproductive years of their life bromocriptine should be used. My own regimen for commencing bromocriptine therapy is to advise the patient to start therapy at night or upon retiring to bed by taking a quarter of a 2.5 mg tablet together with milk and cookies. Every three days the dose is increased by a quarter of a tablet; 1/4 tablet at night and at breakfast; after a further three days at night, at lunch, and on retiring to bed. The dose is increased every 3 days by a quarter of a tablet until the patient takes 2.5 mg with breakfast, at lunch, and on retiring. If the patient develops side effects at any time during the dose escalation they are advised to back off to the previous regimen where they were free of side effects. Many patients are successfully treated with a lower dose than 2.5 mg t.i.d. In my experience it is very unusual for this regimen to fail. Another common error is for the dose to be escalated above 2.5 mg t.i.d. Large tumors are responsive to the same dose of bromocriptine as small tumors. The affinity of the receptor is the same irrespective of the size of the lesion. In the past the larger the tumor and the greater the abnormality in visual fields the greater was the temptation to operate on these patients to save vision. However at the time of diagnosis visual field defects have often been present for years and therefore the establishment of the diagnosis does not constitute a neurosurgical emergency. However, this situation needs to be distinguished from the patient who is absolutely normal and then develops acute pituitary apoplexy, and vision is suddenly lost over the space of minutes, hours, or days. That is a neurosurgical emergency and requires urgent surgical decompression. Macroprolactinomas (diameter > 10 mm) are typically associated with prolactin levels of >250 ug/L. These tumors are cured in <30% of cases in the very best neurosurgical hands. Thus surgery can be effective in debulking these tumors but does not achieve the objectives of restoring prolactin levels to normal, restoring gonadal function and preventing recurrence. Medical therapy is successful in at least 75% of cases, and the only down side of this approach is that treatment needs to be continued for the rest of the life of the patient. In my view a patient with a prolactinoma, irrespective of its size, should be treated medically first. With microadenomas it is less important to decide whether the tumor itself is a prolactinoma, as microadenomas rarely interfere with other pituitary functions and rarely give rise to local symptoms. If the serum prolactin is more than 250 ug/L, it is very likely to be a prolactinoma. However, if the tumor is less than 10 mm in diameter, the prolactin level may be lower, and the tumor could still be a prolactinoma. If the prolactin level is elevated either because of the tumor secreting prolactin or because of interference in the delivery of dopamine to normal lactotrope cells by the presence of the adenoma, medical therapy will lower prolactin levels to normal and restore gonadal function. If the tumor is a prolactinoma, over 75% of such tumors will reduce in size during therapy. This can occur in the space of hours or days. However some tumors may take months to shrink. If visual field defects are present then it is recommended that these should be followed carefully and, in the event of any deterioration, a repeat MRI should be performed to determine whether the tumor has increased in size. Serum prolactin levels should be checked at 6 weeks and at 3, 6, and 12 months after initiation of treatment. If visual field defects are present they should be objectively measured with Goldman or Octopus apparatus and then repeated at the same intervals. If no improvement of vision is observed after 3 months then a repeat MRI should be performed and, if the tumor is still pressing on the optic chiasm, surgical decompression should be considered. In other circumstances, MRI should be repeated at 1 yr to demonstrate size reduction of the tumor. Most patients with macroprolactinomas experience more than 30% reduction in tumor size. Women with microprolactinomas who wish to become pregnant are at very low risk of tumor expansion during pregnancy. However if the macroprolactinoma was compressing the chiasm before medical therapy, then the patient needs to be warned that, when medical treatment is withdrawn, when she is confirmed to be pregnant, the tumor is likely to reexpand to its previous size, and compressive symptoms may develop. The expansion of normal pituitary tissue during pregnancy may exacerbate this problem. At that time medical therapy can be reinitiated or surgical therapy can be considered. Bromocriptine has been used during pregnancy, and no increases in fetal abnormalities over those seen in a normal population have been observed. It does suppress maternal pituitary prolactin secretion and is associated with reduction in pituitary tumor size. It is not approved for this indication in this country. I do not recommend that it be used routinely to prevent compressive symptoms, but if those symptoms develop, then bromocriptine should be seriously considered. Bromocriptine does cross the placenta, and the fetal prolactin secretion is suppressed. Alternatively, prophylactic surgical decompression of the tumor can be considered before pregnancy. It is now 26 yr since bromocriptine was introduced into clinical research. It has been shown to be a very safe medication, and its effects are maintained for as long as it is administered. It is very unusual for patients to escape from its effectiveness in suppressing prolactin levels and reducing tumor size. For long-term management of macroprolactinomas, prolactin levels should be checked on annually, and MRI scans should be repeated if symptoms of tumor expansion occur. Alternatively once shrinkage of tumor size has been documented, the interval between MRI scans can be doubled each time, i.e. at year one, two, four, and eight. It may also be possible to lower the dose of bromocriptine once the prolactin is normal. patients only 2.5 mg but others 2.5 mg t.i.d. patients from although the dose should be as low as possible to achieve of serum prolactin. In medical therapy is the first line of treatment for now have over a quarter of of experience with dopamine agonists, and it is that in the of the of other pituitary tumors will to the of most patients with prolactinomas are initially as well as with bromocriptine or a dopamine the relatively few patients are treated initially by surgical of a prolactinoma, in where a view of surgery In the majority of patients who the drug and hyperprolactinemia will remain on the medication medical management the treatment when surgical treatment is reserved for those patients who are to be medical and in this surgery is with rather than with surgical cure as a likely In my I patients who have been advised to medical but who are unaware of the duration of the for as to and the of side effects. In fact, some patients have that medication would the tumor. In studies have that normal prolactin were in only of patients. that the duration of medical treatment is to the at the time of and that pregnancy may be in the of large medical management less to patients than it does to their The lower published and other to medical therapy for a of the and long-term results of transsphenoidal microsurgery in the initial management of this common tumor. determine the long-term risk of to the results of surgical treatment and to the preoperative and of surgical studied two of women I and treated at the of I was treated more than yr before initiation of the and was treated with a of disease and continued amenorrhea or galactorrhea, no further therapy, and normal prolactin and disease hyperprolactinemia, recurrence of amenorrhea or galactorrhea or further were for purposes of this was in of women in of were in continued after a of lower preoperative prolactin size, and lower and stage were of surgical < for With but not in had a rate initial surgical was in of women who had that were or or prolactin was the of continued < and lower stage were also both initial was in of patients preoperative prolactin were at least of those in the longer in at a of yr after surgery. The men treated in the time were than the women and had larger tumors mm and preoperative prolactin As a they less with disease after surgery and a recurrence rate of patients of the series had with There were no and no patient has had loss of pituitary function in the experience with more than for of days and low morbidity surgical does surgery as the initial treatment make As a the the patient and the the tumor, the the argument for initial surgery. of the sinus is established by a high magnetic resonance imaging scan, most women and men with a prolactin are for surgery the magnetic resonance imaging that I surgery as the primary treatment in any patient the of When is surgery as The patient who medication at an effective dose a surgical surgery may be and if not reduction in the reduces the often lowering the effective dose to a are patients with delayed intolerance after an initial of medical management. commonly patients tumors on a either with or without a rise in prolactin should be treated although may be if is A indication for surgery is the of to shrinkage of a prolactinoma that has into the transsphenoidal are safe and and for many patients the of achieving a clinical for surgery as the initial treatment. recurrence after initial is lower than frequently and in series only after a of in will always but in a some every patient with hyperprolactinemia is treated with medication initially and, I that many patients with tumors of small or size would choose surgical over medical initial treatment if with of present about both medical management the and experience of the and the same to surgical treatment and the