| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Endocrine Care |
Pituitary Unit, Department of Neurosurgery, Istituto Scientifico San Raffaele, Università Vita-Salute, 20132 Milano, Italy
Address all correspondence and requests for reprints to: Marco Losa, M.D., Department of Neurosurgery, Istituto Scientifico San Raffaele, Via Olgettina 60, 20132 Milano, Italy. E-mail: . losa.marco{at}hsr.it
Abstract
Medical therapy with dopaminergic drugs is the preferred initial treatment for symptomatic prolactin (PRL)-secreting adenomas; but in recent years, there has been a renewed interest in surgery. The aim of this study is to report a large series of patients operated for prolactinoma in the last 10 yr. A total of 120 consecutive patients (93 female, 27 male) underwent surgery from January 1990 to December 1999. Their mean age at surgery was 29.7 ± 0.9 yr. Fifty-nine patients (49.2%) had a microadenoma, and the remaining 61 (50.8%) had a macroadenoma, of which 24 (20%) were intrasellar and 37 (30.8%) were extrasellar adenoma. Magnetic resonance imaging signs of invasion of the cavernous sinus were detected in 18 patients (15.0%). Thirty-one patients (25.8%) had never been treated before, whereas the remaining 89 (74.2%) had received dopaminergic drugs. After surgery, normalization of PRL levels occurred in 77 patients (64.2%). Logistic regression analysis showed that the only predictive factor of unsuccessful surgery was a high preoperative PRL level. Recurrence of hyperprolactinemia occurred in 13 of the 77 cured patients (16.9%) during a mean follow-up of 50.2 ± 3.0 months; the 5-yr disease-free survival was 75.9%. Extrasellar extension of the tumor and presence of a postoperative PRL response to TRH were associated with a lower risk of relapse. In summary, surgery normalized PRL levels and relieved symptoms of hyperprolactinemia in most patients. Recurrence of hyperprolactinemia occurred within 4 yr after surgery. Transsphenoidal surgery can be offered as a definitive therapy, especially to patients with intrasellar tumors.
PROLACTIN (PRL)-SECRETING TUMORS represent the most common subtype of pituitary adenoma, usually causing menstrual disturbances and/or galactorrhea in women of reproductive age and loss of libido or impotence in men. Moreover, macroprolactinomas (tumors with a maximum diameter more than 10 mm) may cause symptoms owing to mass effects on surrounding structures. There is consensus that patients with a PRL-secreting adenoma need treatment to reverse hyperprolactinemia and control tumor growth. Only patients with microprolactinomas (tumors with maximal diameter less than 10 mm) who do not present clinically relevant symptoms may be followed up without treatment (1).
Therapy of PRL-secreting adenomas is circumscribed to medical therapy with dopamine agonists and surgery, whereas radiation therapy is recommended to the few patients not controlled by both medical and surgical treatments. Medical therapy reverses clinical symptoms, normalizes PRL levels, and reduces tumor size in most patients with prolactinomas (1, 2). Because of the excellent clinical results with dopamine agonists and a perceived high risk of recurrence of hyperprolactinemia after successful surgery (3), medical therapy has become the preferred initial management of prolactinomas. However, some drawbacks of medical therapy, such as the need for life-long treatment in the majority of patients, side effects limiting compliance by the patient, and resistance of the tumor to dopamine agonists, are still of concern. In recent years, there has been a renewed interest in the surgical treatment of PRL-secreting tumors, especially microprolactinomas (4, 5), which has promoted some controversy about the first-line therapy (6, 7).
The aim of our study is to report the results of a large and homogeneous series of patients operated for a PRL-secreting pituitary adenoma in the last 10 yr. We determined early and late results of surgery, with special emphasis on the risk of and the factors affecting recurrence of hyperprolactinemia.
Materials and Methods
We included, in the study, all the patients undergoing transsphenoidal resection of a PRL-secreting pituitary adenoma at our department from January 1990 through December 1999. All operations were performed by two of us (P.M. and M.G.). The only exclusion criterion was previous pituitary surgery (six patients), because indications and results of repeated-surgery patients may differ considerably from those of patients who have not had previous surgery.
Prospectively recorded data included: age at operation, sex, estimated duration of disease, symptoms at presentation, indication for surgery, history and effectiveness of previous drug therapy, hormonal data, and complications of surgery. Magnetic resonance imaging (MRI) was performed on all patients at the time of diagnosis; and, if they had received dopaminergic drugs, it was usually repeated before surgery. Tumor size and extension were determined on the first MRI. The preoperative serum PRL level used for analysis was that obtained before surgery or, when applicable, the highest value before the start of dopaminergic drugs. Early postoperative PRL levels were measured 5–6 d after surgery. If patients had received dopaminergic drugs or had discontinued them shortly before surgery, the earlier postoperative value considered for classifying surgical outcome and for statistical analysis was that obtained at least 2 months after surgery. TRH test (200 µg iv injected after baseline, with blood samples collected 15, 30, and 60 min thereafter) was performed before and 5–6 d after surgery, unless patients had received dopaminergic drugs until surgery. The criterion for a positive PRL response was an increment of more than 50% over baseline and an absolute rise of at least 5 µg/liter.
PRL concentrations on referral, as well as those during follow-up, were measured at a large number of laboratories that use different assay kits. During the period covered by this study, histopathological diagnosis was based on hematoxylin-eosin-stained sections and immunocytochemical characterization of secretory activity, using commercially available antisera as previously described (8).
Biochemical remission of hyperprolactinemia was defined as postoperative normalization of basal serum PRL levels (<20 µg/liter in women and <15 µg/liter in men) without dopaminergic therapy for at least 2 months. In any case, to avoid misclassification of early surgical results, patients were not considered in remission if hyperprolactinemia recurred within 6 months of surgery. Hypogonadotropic hypogonadism was diagnosed in premenopausal women with amenorrhea and in men with subnormal testosterone levels. Low or normal gonadotropin levels were required in both cases. Secondary hypothyroidism was diagnosed in patients with low free T4 level and normal or suppressed TSH concentration. Secondary hypoadrenalism was diagnosed in patients with low 24-h free urinary cortisol, low morning cortisol level, and/or clinical symptoms of hypoadrenalism responding to replacement therapy with glucocorticoids. Transient diabetes insipidus was diagnosed when hypotonic polyuria (>40 ml/kg body weight daily), usually self-remitting in a few days, ensued soon after surgery; whereas permanent diabetes insipidus was diagnosed when hypotonic polyuria lasted for at least 3 months.
Long-term information of patients not followed at our center was obtained by contacting directly the patient or the patient’s endocrinologist by phone. We collected information (particularly on the menstrual and/or gonadal function, reproductive history, last determination of PRL levels, current or past use of hormonal substitution therapy, and further treatment for hyperprolactinemia) if needed.
Recurrence of hyperprolactinemia was defined as the detection, during follow-up, of elevated PRL levels (usually confirmed in three consecutive blood specimens taken 30 min apart), irrespective of the presence of symptoms and neuroradiological findings.
Statistical analysis
Continuous variables were examined for homogeneity of variance. PRL concentrations, both before and after surgery, were positively skewed. This was corrected by log-transformation before statistical analyses. Continuous data are expressed as mean ± SE, except for PRL levels, which are presented as median values. Correlation coefficients were calculated using regression analysis. Student’s t test for unpaired data was used to compare continuous variables among groups. The Wilcoxon signed-rank test was used to evaluate paired differences of PRL levels before and after surgery. Categorical variables were compared using Pearson’s chi-square test or Fisher’s exact test, as appropriate. Multiple logistic regression analysis was used to determine which variables independently predicted early surgical outcome. We entered in this analysis only those variables that had a P value less than 0.10 in the univariate analysis. The Kaplan-Meier method (9) was used to analyze the primary end point of recurrent hyperprolactinemia during long-term follow-up. Recurrence-free survival was measured from the date of successful surgery to the date of relapse and was censored at the date of the last follow-up. A P value less than 0.05 was considered to indicate statistical significance, and all reported P values are two-sided. All calculations were performed using the statistical package StatView 5.0 (Abacus Concepts Inc., Berkeley, California).
Results
Patient’s characteristics
A total of 120 consecutive patients (93 female, 27 male) underwent surgery for a PRL-secreting pituitary adenoma. Mean age at surgery was 29.7 ± 0.9 yr (range, 12–67 yr). On the basis of MRI findings, 59 patients (49.2%) had a microadenoma, and the remaining 61 (50.8%) had a macroadenoma. In some analyses, the latter group was subdivided into patients with an intrasellar tumor (24 patients, 20.0%) and those with an extrasellar adenoma (37 patients, 30.8%). Invasion of the cavernous sinus, corresponding to grade III-IV according to the classification of Knosp et al. (10), was observed in 18 patients (15.0%), of whom 2 had a microadenoma and 16 a macroadenoma. Thirty-one patients (25.8%) had never been treated before, whereas the remaining 89 (74.2%) had received dopaminergic drugs (usually bromocriptine or, especially in the last years, cabergoline). The duration of treatment varied from 1 month to 15 yr and was intermittent in some cases. Dopaminergic drugs were stopped 2 months or more before surgery in 41 of the 89 medically treated patients. At the time of diagnosis, 105 patients (89.0%) had hypogonadism, 7 (5.9%) had hypothyroidism, and 5 (4.2%) had hypoadrenalism. Indications for pituitary surgery were categorized into 6 groups: resistance to dopamine agonist drugs (failure to suppress PRL levels to greater than half the starting level or, in the case of large tumors, displacing the optic pathway, lack of significant tumor shrinkage despite escalating doses of dopaminergic drugs), drug intolerance, strong personal preference for surgery rather than chronic medical treatment, uncertainty about the correct diagnosis (typically when the patient presented with large and cystic tumors and PRL levels below 200 µg/liter), partial clinical response (when, despite a significant fall of PRL levels during medical treatment, hyperprolactinemia and clinical symptoms persisted), and tumor apoplexy causing sudden neurologic defects. Table 1
summarizes the main indication for surgery according to tumor size. The median preoperative PRL concentration in the total cohort was 114 µg/liter (range, 21–7850). There was a highly significant correlation between maximum tumor diameter and basal PRL concentration (r = 0.777; P < 0.001; Fig. 1). Men, compared with women, had higher preoperative PRL levels (480 µg/liter vs. 88 µg/liter; P < 0.0001), greater maximum tumor diameter (25.8 ± 2.8 mm vs. 11.3 ± 0.7 mm; P < 0.0001), and higher frequency of cavernous sinus invasion (37% vs. 10%; P < 0.01).
|
|
All operations were performed by the transsphenoidal route. Histological analysis revealed a pituitary adenoma with positive staining for PRL in all patients except seven, whose histology result was negative. All these seven patients had a microadenoma and PRL levels ranging from 34–160 µg/liter.
No operative or perioperative mortality occurred. Major morbidity occurred in six patients (5%) and consisted of permanent visual damage in one eye in a patient with a large and invasive macroadenoma, transient worsening of vision in one eye in another patient with an extrasellar macroadenoma, deep vein thrombosis requiring anticoagulant therapy for 1 yr in one patient with an intrasellar macroadenoma, epistaxis requiring emergency nasal tamponade 10 d after surgery in one patient with a microadenoma, mucocele requiring evacuation 1 yr later in one patient with a microadenoma, and renal colic on the fourth postoperative day in one patient with a microadenoma.
Thyroid and adrenal deficiency of new onset occurred in only 2 patients of 112 and 115 at risk, respectively. Patients with postoperative impairment all had an extrasellar macroadenoma. Therefore, no new deficit of anterior pituitary function occurred in patients with microadenomas and intrasellar macroadenomas. On the other hand, only 1 patient of 7 recovered impaired thyroid function, and none out of 5 recovered impaired adrenal function. Hypogonadism did not develop in the 13 patients with normal gonadal function at presentation. Transient diabetes insipidus occurred in 52 patients (43.3%), whereas permanent diabetes insipidus occurred in 8 patients (6.7%), of whom 5 had an intrasellar adenoma and the remaining 3 an extrasellar tumor. Six patients became asymptomatic and no longer required desmopressin therapy by 10 months; whereas, in the remaining 2 patients, diabetes insipidus was still present 32 and 48 months after surgery, respectively.
Most patients experienced a marked improvement in symptoms. Sixty-seven of the 79 premenopausal women with preoperative amenorrhea (84.8%) regained normal menses after surgery. Only one woman had persistent amenorrhea despite normalization of PRL levels. On the contrary, only 7 of the 18 women who did not attain normal PRL levels after surgery regained normal menses. In most of the latter patients, PRL levels declined near the upper normal limit. In men, restoration of normal sexual function was reported in only 5 of the 13 patients (38.5%) with preoperative loss of libido. During follow-up, 28 female patients spontaneously conceived (26 were in the group of postoperative remission, whereas the remaining 2 were still hyperprolactinemic after surgery and were further treated with dopaminergic drugs). Nine other patients reported infertility (3 were cured, whereas the remaining 6 were either not cured or had a recurrence of the disease). The other 56 female patients did not wish to conceive. Information on the presence of headache before surgery was available for 111 patients, of which 55 patients (49.5%) reported mild or severe headache. At the first postoperative examination, 50 of the 55 patients with headache reported no (30 cases) or only slight headache (20 cases). Interestingly, improvement of headache occurred to the same extent in both cured and not-cured patients.
The median postoperative PRL concentration in the whole group of patients was 7.8 µg/liter (range, 0.3–2680 µg/liter). This value was significantly lower than the preoperative one (P < 0.001). Surgical cure, defined according to the criteria outlined in Material and Methods, was achieved in 77 patients (64.2%). Normalization of PRL levels occurred in 46 of the 59 patients with microadenoma (78.0%) and in 31 of the 61 patients with macroadenoma (50.8%; P < 0.01). In the latter group, patients with intrasellar adenoma had a higher rate of surgical cure than patients with extrasellar tumor (87.5% vs. 27.0%; P < 0.01). Because of the similar rates of cure, patients with microadenoma and intrasellar adenoma were grouped together in all subsequent analyses. Univariate analysis revealed that higher preoperative PRL levels, larger tumor size, male sex, younger age at surgery, invasiveness into the cavernous sinus, and dural infiltration were associated with a negative surgical outcome (Table 2); whereas PRL responsiveness to TRH, positive histology, and resistance to dopaminergic drugs were not associated with surgical outcome (Table 2). The same holds true for a history of previous dopaminergic therapy (Table 2), even when the data are analyzed according to tumor size. Indeed, 17 of 20 patients with intrasellar tumors not treated with dopaminergic drugs (85%) had a good surgical outcome, compared with 50 of 63 previously treated patients (79.4%); whereas 7 of 26 patients with extrasellar adenomas not treated with dopaminergic drugs (26.9%) were in remission, compared with 3 of 11 previously untreated patients (27.3%).
|
), we next performed a multiple logistic regression analysis. In the first model, we included (as a dependent variable) surgical outcome and (as independent variables) age, sex, tumor size, preoperative PRL level, invasiveness into the cavernous sinus, and resistance to dopaminergic drugs. The only variable associated with unsuccessful surgery was a high preoperative PRL level, whereas all other variables had no independent association with early outcome. In the second model, we included all the aforementioned variables plus histological infiltration of the dura. Because of missing values of the latter variable, this analysis included only 50 patients; but, again, preoperative PRL levels emerged as the only independent variable associated with surgical outcome. Long-term results
Mean follow-up for the whole group of patients was 50.2 ± 3.0 months, ranging from 1–132 months. Nine patients could not be traced by phone, despite repeat attempts, and were considered lost to follow-up. In this group, six patients were cured and three not cured. The last follow-up information of these patients was obtained 6, 6, 10, 12, 22, and 24 months after surgery among cured patients and 1, 8, and 15 months after surgery among not-cured patients.
Recurrence of hyperprolactinemia occurred in 13 of the 77 patients (16.9%) with normal postoperative PRL levels. Fig. 2 shows the disease-free survival curve in our patients. Almost all recurrences occurred within the first 3 yr after surgery, whereas no relapse was recorded 4 yr after surgery. The disease-free interval at 5 yr was 75.9% (95% confidence interval, 63.7–88.1%). Several clinical characteristics, such as sex, age, previous dopaminergic therapy, reason for surgery, preoperative PRL levels (categorized according to the median value), preoperative responsiveness to TRH, histological confirmation of adenoma, and dural infiltration, were not associated with a higher risk of relapse. On the other hand, the 10 tumors with extrasellar extension did not recur, as compared with 13 relapses in the group of the remaining 67 microadenomas and intrasellar macroadenomas. The only other characteristic associated with a lower risk of relapse was the presence of a positive PRL response to TRH in the immediate postoperative period (Fig. 3). In this analysis, we excluded 10 patients with a preoperative positive PRL response to TRH and 8 patients who had not performed the test either before or after operation. PRL levels after surgery did not affect the risk of recurrence. We used 2 different PRL cut-offs: the first was 4 µg/liter, corresponding to the median PRL value in the 77 patients considered in remission; and the second was 10 µg/liter, an arbitrary level chosen to analyze separately patients with PRL values in the upper normal range. In both cases, the log-rank test was not significant (P = 0.19 and P = 0.56, respectively).
|
|
summarizes the treatment received by the 43 patients not cured after surgery and the 13 patients who relapsed during follow-up. Most of the patients who received dopaminergic drugs alone (either bromocriptine or cabergoline) had normalization of PRL secretion. Medical plus radiosurgical treatment also led to normalization of PRL secretion in most patients who received both treatments. Notably, all the patients treated by the 
-knife were resistant or partially responsive to dopaminergic therapy and did not attain normal PRL levels with medical treatment alone. Three patients required repeat surgery because of tumor regrowth despite continued dopaminergic treatment. No further treatment was advised in 6 patients because, despite slightly elevated PRL levels, they had no symptoms attributable to hyperprolactinemia. In 1 of these 6 patients, normalization of PRL levels occurred spontaneously, 1 yr after surgery.
|
Discussion
Since the introduction in clinical practice of dopamine agonists more than 25 yr ago, indications for surgery in patients with PRL-secreting adenomas have become less frequent. Indeed, during the period covered in this study, PRL-secreting tumors accounted for 14.7% of all patients who received their first surgery in our department; whereas, in the period 1970–1979, they represented 31.6% of all pituitary adenomas operated by one of us (M.G.). Our current series, therefore, reflects the indications and results of surgery in the era of dopamine agonists (Table 1). The relatively high number of patients resistant or intolerant to dopamine agonists is likely to reflect a referral bias to a center specialized in pituitary surgery.
Overall, the early surgical remission rate was 64.2%. Females and patients with intrasellar tumors had higher rates of remission (75.2% and 80.7%, respectively). These figures compare well with most of the other published series (3, 5 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24). Improvement of clinical symptoms occurred in almost all patients with normalization of PRL levels and even in some patients with persistent hyperprolactinemia after surgery. This was especially true for headache. Only sexual disturbance in men was less likely to regress after surgery. The most plausible explanation for this finding is the higher frequency of extrasellar and invasive adenomas in men and the consequent greater probability of irreversible pituitary damage. About 90% of the women complaining of infertility at presentation could conceive after surgical normalization of hyperprolactinemia, underlying the good preservation of pituitary function after surgery. Restoration of fertility has been already reported in other series (3, 5, 11, 16, 18, 22, 25, 26). However, few studies correlated the number of women conceiving with those wishing to become pregnant. In 1 study, 11 of 13 women (85%) wishing to conceive became pregnant after surgical normalization of hyperprolactinemia (19). Similar high rates of pregnancy (90% and 88.3%) were reported in other studies (25, 26), whereas slightly lower pregnancy rates (75%) occurred in other 2 series (12, 13). In the latter studies, all the women wishing to conceive were included, independently of the postoperative PRL levels. Indeed, Woosley et al. (13) concluded that infertility was attributable to persistent hyperprolactinemia.
Successful surgical outcome was associated with several clinical characteristics (Table 2). Most of them, such as low PRL basal levels and small tumor size, have already been recognized in previous surgical series (11, 14, 16, 20, 21, 23, 24). However, with the exception of the series by Tyrrell and co-workers (24), no study employed multivariate analysis methods to investigate the independent role of the different variables. The relevance of this point is underscored by the fact that most variables significantly associated with early outcome in the univariate analysis (e.g. tumor size, sex, age, cavernous sinus invasiveness, and dural infiltration) lost their independent value in the multivariate analysis. In our analysis, PRL level was the only predictive variable. Tyrrell and co-workers reported that both lower preoperative PRL levels and lower adenoma stage were the best predictors of immediate surgical outcome. In contrast to these data, Turner and co-workers (5) found no significant difference of preoperative PRL levels among 25 cured and 7 not-cured patients. The most plausible explanation for the discordant results is that Turner and co-workers (5) included only patients with microprolactinoma.
Previous therapy with dopaminergic drugs did not affect surgical results, independently of tumor size at presentation. The question whether medical treatment with dopamine agonists may jeopardize surgical outcome has been debated since the first report by Landolt et al. (27) in 1982. They found that in patients with microprolactinomas, the surgical cure rate fell from 81% in untreated patients to 33% in medically treated patients. The negative effect of medical pretreatment was ascribed to induction of perivascular fibrosis by bromocriptine (28), which might render surgical removal of the tumor more difficult. However, subsequent studies did not confirm this experience in microprolactinomas (5, 21, 24, 29, 30, 31). Most studies reported worse results in macroprolactinomas previously treated with dopamine agonists (19, 21), but it is unclear whether the negative outcome was directly caused by the drug or whether it was attributable to an increased tendency to treat large and invasive tumors, before surgery, with dopamine agonists.
Recurrence of hyperprolactinemia is one of the major caveats of surgery for PRL-secreting adenomas. The very wide variation in reported rates of relapse, ranging from 0–55% (Table 4), is one of the most controversial aspects in the field of pituitary surgery. It is possible that lack of strict hormonal criteria to define remission of hyperprolactinemia, such as that proposed in acromegaly or Cushing’s disease, may cause some misclassification in the early postoperative period, even though all studies relied on the same criterion to define surgical cure, i.e. normalization of basal PRL levels.
|
). It is possible that the personal surgeon’s choice between the less aggressive selective removal of the tumor or the more aggressive partial hypophysectomy with excision of a generous rim of surrounding tissue may explain some of the variability in the reported rates of recurrence. However, different surgical techniques did not affect early or late outcome in patients with Cushing’s disease (33). Disease-free survival analysis allowed us to estimate that the recurrence rate in our patients is 24.1% at 5 yr (95% confidence interval, 11.9–36.3%). No direct comparison can be made with other studies, because they did not use survival curve analysis. All the recurrences occurred within 4 yr after surgery. Time to recurrence has varied widely among the different surgical series. In some studies, the median time to relapse was longer than 5 yr (24, 25); whereas in others, almost all relapses occurred in the first 3 yr after surgery (3, 15, 18, 21, 22). Some studies looked specifically at detecting risk factors for recurrent hyperprolactinemia (Table 4). A higher postoperative PRL level was identified as an unfavorable prognostic factor in most studies. In our experience, basal PRL levels measured 5–6 d after surgery, or within 2 months in those patients taking dopaminergic drugs until surgery, did not predict late outcome. One possible explanation for the diverging results is the exclusion from the category of recurrence of those patients with elevated PRL levels within 6 months of surgery, which we classified directly as surgical failures. Another possible explanation is methodological, in that we used survival curves to analyze recurrence of hyperprolactinemia; whereas all other studies, with 1 exception (24), used less appropriate statistical analysis. Despite the low number of patients available for analysis, early postoperative absence of PRL response to TRH administration identified patients at higher risk of subsequent relapse. Abnormal regulation of PRL secretion by TRH could be ascribed to persistence of tumorous lactotroph cells. In our study, 20 patients who resumed a normal PRL rise to TRH after surgery had no recurrence during follow-up, whereas relapse occurred in 9 of the 39 patients with absent PRL response to TRH. However, it should be noted that not all patients without the PRL response to TRH developed recurrence of hyperprolactinemia, underscoring the low specificity of this finding. Similar results have been reported by others (18, 20). The percentage of patients resuming a PRL response to TRH was lower in our study than in previous studies (18, 20), probably because we tested our patients within 1 week from surgery, whereas patients were tested 6 months after surgery in 1 study (18). It is likely that soon after removal of the PRL-secreting adenoma, normal lactotrophs may still be functionally suppressed and, hence, not be able to respond to TRH. Anyway, the low specificity of the absent PRL response to TRH to predict recurrence of hyperprolactinemia seems a more general problem, given that, even in the study by Schleckte et al. (18), 12 of 17 patients who had absent PRL response to TRH 6 months after surgery did not develop a recurrence. No association between tumor size and the risk of relapse has been found in most studies (Table 4), even though, in one series, recurrence occurred in 31% of patients with microadenoma and in 91% of those with macroadenoma (32). Another study found a higher risk of relapse in tumors of higher stage, i.e. invasiveness, but no association with tumor size or grade (24). Our results are clearly different, because we detected no relapse in the 10 extrasellar tumors, as compared with 13 recurrences among the 67 intrasellar tumors. Our data should be interpreted cautiously, however, because of the small number of patients with extrasellar tumors at risk of relapse. Moreover, 5 of the 10 patients with extrasellar tumors initially cured by surgery had a cystic adenoma with relatively low preoperative PRL levels. It is possible that such tumors, which had already undergone necrotic-hemorrhagic changes, may have a better outcome than solid adenomas of similar size and extension.
Most patients with either persistent or recurrent hyperprolactinemia received further treatment. Six patients were not further treated because, despite abnormal PRL levels, they had no symptoms attributable to hyperprolactinemia. Stable clinical improvement after incomplete tumor removal has been already described in other series (5, 12, 16, 21); and, in such cases, biological cure of hyperprolactinemia has been achieved. Further treatments were successful in 72.9% of patients. Good results were obtained with -knife radiosurgery. Indeed, four of the seven evaluable patients normalized PRL levels after radiosurgery. Considering that all patients had residual tumors invading the cavernous sinus and were resistant to dopaminergic drugs, this is a very interesting result. A recent report confirms the efficacy of -knife in selected patients with PRL-secreting adenomas previously operated on (34).
Surgery was safe in our patients. No perioperative mortality occurred, and major morbidity was confined to 6 patients, of which only 1 (with a giant and invasive tumor) had a permanent neurologic deficit. Impairment of pituitary function occurred infrequently, especially in intrasellar adenomas, and truly permanent diabetes insipidus was observed in only 2 patients. These results seem slightly better than those of 31 pooled surgical series including almost 2500 patients (35) but similar to those of major series at a single center experienced in pituitary surgery (20, 21, 24, 26).
We conclude that, besides the already accepted indications of intolerance or resistance to medical therapy, transsphenoidal surgery should be offered to patients with newly diagnosed intrasellar PRL-secreting tumors, whereas it should be avoided in the case of extrasellar extending tumors with high preoperative PRL levels. As suggested by others (24), careful selection of patients is of uttermost importance, to maximize surgical cure rates while diminishing side effects. Experience of the surgeon is important as well, and each surgical team should assess its own results for efficacy and safety (5). Surgery can obviate the need for chronic drug treatment and, in the long run, might be more cost-effective than medical therapy (5). Our current approach to patients with PRL-secreting adenomas is to advise drug therapy in the case of extrasellar macroadenomas and to discuss openly and objectively with the patient the advantages and disadvantages of either medical and surgical treatment in the case of intrasellar tumors, allowing the patient to decide what is the best approach in her or his situation.
Footnotes
Abbreviations: MRI, Magnetic resonance imaging; PRL, prolactin.
Received December 4, 2001.
Accepted March 21, 2002.
References
This article has been cited by other articles:
 |
|
 |
|
  J Kreutzer, R Buslei, H Wallaschofski, B Hofmann, C Nimsky, R Fahlbusch, and M Buchfelder Operative treatment of prolactinomas: indications and results in a current consecutive series of 212 patients Eur. J. Endocrinol., January 1, 2008; 158(1): 11 - 18. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. P. Gillam, M. E. Molitch, G. Lombardi, and A. Colao Advances in the Treatment of Prolactinomas Endocr. Rev., August 1, 2006; 27(5): 485 - 534. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Colao, A. Di Sarno, P. Cappabianca, C. Di Somma, R. Pivonello, and G. Lombardi Withdrawal of Long-Term Cabergoline Therapy for Tumoral and Nontumoral Hyperprolactinemia N. Engl. J. Med., November 20, 2003; 349(21): 2023 - 2033. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. A. Schlechte Prolactinoma N. Engl. J. Med., November 20, 2003; 349(21): 2035 - 2041. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
