Search

To determine if the mode of 17 beta-estradiol (E2) administration affects growth hormone (GH) concentrations, eight postmenopausal women were studied under the following conditions: (1) control (no E2), (2) oral E2 (Estrace, 1 mg every 12 h for 2 weeks) and (3) transdermal E2 (Estraderm patch, 0.1 mg, two patches changed daily for 2 weeks). Blood was collected every 5 min for 24 h and assayed for serum GH concentrations using a sensitive chemiluminescence assay. Serum E2 levels were comparable during both E2 treatment regimens when measured with a specific chemiluminescence assay. The 24-h integrated GH concentrations (IGHC, min . micrograms/L) increased in all eight subjects from (mean +/- SE) 494 +/- 102 during control to 860 +/- 111 (P0.05) and 832 +/- 149 (P0.05) during oral and transdermal E2, respectively. Both E2 treatments significantly increased GH pulse height, individual pulse area, incremental pulse amplitude, interpeak valley concentration, and interpeak valley nadir (as measured by Cluster algorithm) when compared with control. No significant differences were observed in the number of GH pulses per 24 h. Insulin-like growth factor-I (IGF-I, micrograms/L) concentrations decreased from 165 +/- 19 (control) to 109 +/- 11 (oral E2, P0.05) and 122 +/- 15 (transdermal E2, P0.05). No statistically significant differences in attributes of pulsatile GH release or IGF-I concentrations were observed between the oral and transdermal E2 treatments. We conclude that both oral and transdermal E2 treatment increase serum GH concentrations in postmenopausal women. This increase is manifested by larger GH pulses and higher basal (interpulse) GH levels, not by changes in pulse frequency. Both routes of E2 administration decrease serum IGF-I concentrations, which may attenuate IGF-I negative feedback on pituitary somatotrophs and hypothalamic somatostatin secretion, resulting in enhanced pulsatile GH release.

Thirty-four human pituitary adenomas were examined for the presence of Pit-1 mRNA via Northern analysis. All tumors that were immunoreactive with either human (h) PRL or hGH antiserum contained Pit-1 mRNA (8 of 8) as did a small percentage of alpha-subunit only (1 of 3) and null tumors (1 of 9). Two tumors that contained TSH beta mRNA by RNAse protection assay also contained Pit-1 mRNA. Tumors that were immunoreactive with antiserum for hACTH, hFSH beta, or hLH beta without containing PRL, GH, or TSH beta were uniformly Pit-1 negative (0 of 14). Although RNA from the majority of Pit-1-positive tumors (10 of 11) contained primarily a 2.4-kilobase (kb) Pit-1 mRNA transcript, 1 tumor appeared to contain a 1.6-kb mRNA. Analysis of human pituitary autopsy specimens revealed no 1.6-kb Pit-1 mRNA, indicating that this variant may be tumor specific. The wide variety of serum PRL, GH, and insulin-like growth factor-I levels in patients with Pit-1-positive tumors suggests that although Pit-1 is necessary for PRL and GH production, it is not uniformly associated with hormone production, as determined by either immunohistochemistry or elevated serum levels. The expression of Pit-1 mRNA in a small fraction of null and alpha-subunit only tumors indicates that there may be a previously unsuspected diversity of origin within these particular subsets of adenomas.

Forty-one human pituitary adenoma specimens were examined for the presence of estrogen receptor (ER) messenger ribonucleic acid and protein using a combination of ribonuclease protection assay, [3H] estradiol ([3H]E2) binding, and ER immunohistochemistry. ER messenger ribonucleic acid prevalence was high in PRL-immunoreactive tumors (2 of 2), moderate in GH/PRL tumors (2 of 5), and low or absent (0 of 4) in GH tumors. In the GH/PRL-immunostaining tumors, the presence of the ER was uniformly associated with elevated serum PRL levels. Among the gonadotropin-immunostaining tumors, 10 of 17 were ER positive; within this group, those with gonadotroph adenoma characteristics were ER positive, whereas those with null cell/oncocytic characteristics were ER negative. Of the tumors that did not immunostain for any known anterior pituitary hormones, 3 of 11 were ER positive. ER immunohistochemistry in 14 tumors revealed a 100% correlation with ribonuclease protection assay results, whereas [3H]E2 binding, determined in 9 tumors, showed an 87% correlation. In summary, it appears that PRL and a specific class of gonadotropin-immunostaining tumors (identifiable by specific characteristics on electron microscope) contain ER, whereas GH-immunostaining tumors are ER negative. ER expression in normal pituitary paralleled that in macroadenomas (GH, 2.3%; PRL, 50%; FSH, 70%; LH, 83%; TSH, 4%; ACTH, 1%). The ER-positive tumors represent a subset whose growth and secretory profiles may be influenced by the gonadal steroidal milieu or by pharmacological agents that affect E2 levels or ER function.