This report describes the results of a long term prospective study of 30 women with hyperprolactinemia who were not treated and who underwent yearly clinical, hormonal, and radiographic evaluation for an average of 5.2 yr (range 3-7 yr). At entry into the study 18 women had amenorrhea, 8 had oligomenorrhea, and 4 had regular menstrual periods. The initial mean serum PRL levels did not differ in women grouped according to menstrual function. Nine women (35%) had improvement in clinical symptoms. Serum PRL decreased, and menstrual periods normalized more often in those who initially had oligomenorrhea or regular menstrual periods. In most amenorrheic women serum PRL levels did not decline, and menstrual symptoms did not improve. Six of 30 women had an increase in serum PRL, 14 had no change, and 10 had a decrease, in 6 of whom serum PRL was normal at the last observation. Twenty-seven women had serial radiographic studies. Four (15%) of the 13 women with initially abnormal radiographic findings had normal studies later, 2 had tumor progression, and 7 no change. Four of 14 women who had normal radiographic studies initially developed radiographic evidence of a pituitary tumor, although the radiographic changes were minimal, and no patient developed a macroadenoma or pituitary hypofunction. Increases or decreases in serum PRL did not accurately predict changes in tumor size. Prior estrogen use and previous pregnancies did not increase the likelihood of tumor appearance or enhance tumor growth. The clinical presentation of the patient was an important factor in predicting which patients had a decline in serum PRL and resolution of symptoms. We conclude that patients with hyperprolactinemia are unlikely to have progression of their disease and may, in fact, have clinical and radiographic improvement.

A B S T R A C T Daily administration of estrogen antagonists to premenopausal women has been incorporated into the adjuvant treatment ofbreast cancer. We have studied the changes in reproductive hormones, pituitary responses to hypothalamic-releasing hormones, and endometrial histology during treatment with the antiestrogen tamoxifen in five healthy, premenopausal women. These studies were carried out during one menstrual cycle before and during two cycles ofantiestrogen treatment. All subjects continued to have regular menses with biphasic basal body temperature records. During treatment, estradiol (E2) levels were increased but followed the usual pattern reflecting follicular maturation and corpus luteum formation. The mean E2 concentration at the midcycle peak and during the luteal phase was twice that observed during the non-treatment cycle. By contrast, the concentrations and secretory patterns of luteinizing hormone and follicle-stimulating hormone were not greatly changed, and the gonadotropin responses to gonadotropin-releasing hormone were not suppressed.Endometrial biopsies obtained during the follicular phase of control and tamoxifen treatment cycles showed no differences whereas biopsies obtained during the luteal phase of tamoxifen cycles uniformly showed a lack of changes attributed to progesterone action with no progression of histologic changes beyond those expected on day 7-8 of the luteal phase.These observations are consistent with maturation of multiple ovarian follicles, a surprising finding considering the normal gonadotropin concentrations. The retarded development of the endometrium in the presence of supranormal serum E2 and progesterone concentrations is a morphologic demonstration of the antiprogestational effect of antiestrogens. The lack of gonadotropin suppression in the presence of hyper-

No abstract

To examine the long term effectiveness of transsphenoidal microsurgery for patients with PRL-secreting pituitary tumors, we studied 54 women at yearly intervals after transsphenoidal surgery. Five years after surgery, 19 women (35%) had normal serum PRL concentrations, and 23 (43%) had persistent hyperprolactinemia. Hyperprolactinemia recurred in 12 of 31 patients (39%) who had normal PRL concentration 6 weeks after surgery. None of the patients with recurrent hyperprolactinemia had radiographic evidence of tumor regrowth, and only 3 of 12 had amenorrhea. A serum PRL level below 6 ng/ml 6 weeks after surgery occurred more frequently in cured patients than in those who had a recurrence. PRL responses to TRH were normal in cured patients 1 and 5 yr after surgery and abnormal in those who had recurrent hyperprolactinemia. The PRL responses to chlorpromazine- and insulin-induced hypoglycemia were blunted in patients with normal as well as elevated PRL levels. Patients with recurrent, as well as those with persistent, hyperprolactinemia had no nocturnal rise in serum PRL 5 yr after surgery. The 39% recurrence rate of hyperprolactinemia and persistent abnormalities in pituitary-hypothalamic regulation of PRL secretion after transsphenoidal surgery raise important questions about the choice of primary therapy for patients with PRL-secreting tumors.

No abstract