Pituitary Function in a Man with Congenital Aromatase Deficiency: Effect of Different Doses of Transdermal E2 on Basal and Stimulated Pituitary Hormones

To clarify the role of estrogen on male pituitary function, the effects of different doses of transdermal E2 on pituitary secretion were evaluated in a man with aromatase deficiency. The study protocol was divided into the following three phases: no E2 treatment (phase 1); 25 μg transdermal E2 twice weekly for 9 months (phase 2);12.5 μg transdermal E2 twice weekly for 9 months (phase 3). Pituitary function was studied in detail during each phase of the study protocol by measuring hormone levels in basal conditions and after dynamic testing (GnRH, insulin tolerance test, GHRH plus arginine, TRH, and corticotropin-releasing factor; tests). Basal and GnRH-stimulated gonadotropin levels resulted inversely related to E2 serum levels, according to the dosage of estrogen administered. Basal and stimulated GH, PRL, and TSH serum levels did not change during the protocol study. The secretory pituitary reserve of GH was clearly impaired. Basal and stimulated ACTH and cortisol serum levels were not modified by estrogen administration. This study demonstrated that in the human male E2 is required at pituitary level for normal functioning of gonadotropin feedback both in basal and stimulated conditions. In this patient GH deficiency seems to be an adult-onset event since he reached a tall stature. However, the finding of a severe impairment in GH response to potent provocative stimuli together with the insensitivity of GH/IGF-I axis to circulating estrogens strongly suggest a possible involvement of estrogens on both the development and maturation of the somatotrophic axis. Finally, the congenital lack of estrogen activity seems to be associated with a slightly impaired secretion of PRL and TSH, suggesting a possible role of estrogens on the pituitary secretion of these hormones in the human male.

THE ROLE OF ESTROGENS on pituitary function in the human male is not completely ascertained. Data available from studies performed on young men affected with congenital estrogen deficiency (1–5) and transgenic mice lacking estrogen activity, i.e. α-estrogen receptor knockout mice; β-estrogen receptor knockout mice; and aromatase knockout mice (6), suggest that estrogens play a pivotal role on gonadotropin feedback modulation in the male, acting at both pituitary (7, 8) and hypothalamic level (9). In fact, despite normal or elevated serum T levels, men with estrogen resistance as well as those with aromatase deficiency showed gonadotropin levels above or at the upper limit of the normal range (1–3). Indeed, gonadotropins significantly decreased during estrogen treatment in a dose-dependent manner in aromatase-deficient males (3–5). Conversely, the impact of estrogen on the secretion of pituitary hormones other than gonadotropins is still a matter of debate. In particular, it is well established that estrogens stimulate GH/IGF-I axis in women as androgens do in men (10, 11), probably accounting for sexual dimorphism of this axis at puberty (12–15). However, a possible role for estrogens on GH secretion has been proposed also for men (12). Similarly, modulation of TSH and PRL secretion by estrogens has been suggested based on the observation that the secretion of these two hormones is dimorphic and is increased by estrogen administration in normal men (16, 17). Conversely, ACTH secretion seems to be not affected by estrogens.

The present study was designed to establish the role of estrogens on pituitary function in an adult male with aromatase deficiency. In fact, aromatase deficiency represents a unique natural model to study the effects of both congenital estrogen deprivation and estrogen treatment on pituitary function in humans.

Subjects and Methods

Subject

The study was performed on a male subject with severe aromatase deficiency, who has been previously reported (3). When the diagnosis of aromatase deficiency was confirmed, estrogen replacement therapy was started with a high dose of transdermal E2 (Estraderm, Novartis Pharmaceuticals, Basel, Switzerland; 50 μg twice weekly) in the attempt to obtain epiphyseal closure. After 6 months when the epiphyseal closure was achieved, the patient was put into a different estrogen regimen, according to the study protocol. At the beginning of the study, the patient was 39 yr old. The patient was 189 cm tall, and he weighed 101 kg with a body mass index of 28.3 kg/m² .

Study protocol

To establish the effects of different doses of transdermal E2 on pituitary function, the patient, after giving his informed consent to the treatment, was reassessed after a washout period of 3 months from estrogen treatment (phase 1) and after two subsequent periods of 9 months of transdermal E2 treatment (Estraderm, patch system) at the dose of 25 μg twice weekly (0.47 μg/kg weekly; phase 2) and at the dose of 12.5 μg twice weekly (0.23 μg/kg weekly; phase 3).

Assessment of pituitary function

After the end of each phase of the study protocol, basal serum T, E2, inhibin B, LH bioactivity, FSH bioactivity, SHBG, IGF-I, free T₄ (FT₄) and free T₃ (FT₃) were evaluated and the following test substances applied at 0800 h after an overnight fast: GnRH (100 μg iv Gonadorelin, Ferring Pharmaceuticals Ltd., Kiel, Germany) with blood sampling at −15, 0, 30, 45, 60, and 90 min for LH, FSH, and α-subunit measurement; insulin tolerance test (ITT) (0.1 U/kg iv of insulin) with blood sampling at 0, 15, 30, 45, 60, and 90 min for GH and glucose, PRL, ACTH, and cortisol determination; GHRH (1 μg/kg iv GHRH-29, GEREF, Serono, Italy) plus arginine (ARG) infusion (l-arginine hydrochloride, 0.5 g/kg iv over 30 min from 0 to 30 min) with blood sampling at −15, 0, 15, 30, 60, 90, 120, and 180 min for GH determinations; TRH test (200 μg iv Protirelin, Ferring Pharmaceuticals Ltd.) with blood sampling at −15, 0, 15, 30, 45, 60, and 120 min. for TSH, α subunit, and PRL; corticotropin-releasing factor (CRF) (100 μg hCRF, C-110, CLINALFA AG, Läufelfingen, Switzerland) with blood sampling at −15, 0, 15, 30, 45, and 60 min for cortisol and ACTH determination. Serum and plasma samples obtained from blood collection were stored at −80 C until assays. For every phase of the study protocol, the patient was admitted to the Endocrine Unit for 5 d to execute pituitary dynamic tests separately. Accordingly, only one dynamic test was performed in a single day.

Hormonal assay

Serum total T was measured by commercial RIA (Diagnostic Products, Los Angeles, CA). The inter- and the intraassay coefficients of variation for T were 11% and 5%, respectively.

Serum E2 was detected by a third-generation RIA (Diagnostic Systems Laboratories, Inc., Webster, TX) with a sensitivity of 2.2 pmol/liter. The inter- and the intraassay coefficients of variation for E2 were 4.1% and 3.5%, respectively.

Serum SHBG was measured by a chemiluminescent immunometric assay (Immulite, Diagnostic Products). The inter- and the intraassay coefficients of variation for SHBG were 5.8% and 4.1%, respectively.

Serum LH was measured by a fluoroimmunoassay (Autodelfia hLH kit, Wallac, Inc., Turku, Finland) with a sensitivity of 0.05 IU/liter. The inter- and intraassay coefficients of variation were 3.6% and 2.8%, respectively. LH bioactivity was measured by means of a previously described mouse Leydig cells in vitro bioassay method (18).

Serum FSH was measured by a fluoroimmunoassay (Autodelfia hFSH kit, Wallac, Inc.) with a sensitivity of 0.05 IU/liter. The inter- and intraassay coefficients of variation were 4.1% and 2.6%, respectively. FSH bioactivity was measured by means of a previously described rat Sertoli cell aromatase assay (19).

Serum inhibin B was measured by a two-site ELISA (Serotec, Kidlington UK) with a sensitivity of 15 pg/ml. The inter- and intraassay coefficients of variations were less than 10%. The cross-reactivity with inhibin A was 0.85.

α Subunit circulating pituitary glycoprotein hormone α-subunit was evaluated by an immunoradiometric, noncompetitive, solid-phase assay (Biocode, Liège, Belgium) with a sensitivity of 0.02 mg/liter. The inter- and intraassay coefficients of variations were less than 9% and less than 8%, respectively (20).

Serum GH was measured by a fluoroimmunoassay (Autodelfia hGH kit, Wallac, Inc.) with a sensitivity of 0.01 ng/ml. The inter- and intraassay coefficients of variation for GH were 5.5% and 4.9%, respectively.

Serum IGF-I concentrations were measured by a commercial RIA kit (INCSTAR Corp., Stillwater, MN) after removal of binding proteins by acidification and filtration on octadecyl sulfate C18 cartridges, with a detectability level of 0.02 ng/ml and inter- and intraassay coefficients of variation of 12% and 8%, respectively.

Serum PRL was measured by a fluoroimmunoassay (Autodelfia hPRL kit, Wallac, Inc.) with a sensitivity of 0.04 ng/ml. The inter- and intraassay coefficients of variation for PRL were 3.5% and 2.9%, respectively.

Serum TSH was measured by a fluoroimmunoassay (Autodelfia hTSH kit, Wallac, Inc.) with a sensitivity of 0.03 mU/liter. The inter- and intraassay coefficients of variation for PRL were 3.5% and 2.9%, respectively.

FT₄ and FT₃ were measured by a fluoroimmunoassay (Autodelfia, Wallac, Inc.) with a sensitivity of 2 and 1 pmol/liter, respectively. The inter- and intraassay coefficients of variation were less than 3.1% in both assays.

Serum cortisol levels were measured by RIA methods (Diagnostic Products) with a detectability level of 10 nmol/liter and inter- and intraassays coefficients of variation of 9% and 4%, respectively.

Plasma ACTH was measured by a chemiluminescent immunometric assay (Immulite, Diagnostic Products) with a sensitivity of 0.2 pmol/liter. The inter- and intraassay coefficients of variation for ACTH were 8.9% and 6.1%, respectively.

Results

Serum E2 levels increased during transdermal E2 treatment reaching highest values during phase 2, accordingly to the dose administered (Table 1). Estrogen replacement therapy resulted in the normalization of circulating E2, reaching values within the normal range for adult men during both phases 2 and 3. No side effects, notably gynecomastia and loss of libido, were observed throughout the whole study protocol. No side effect was noted during dynamic testing.

Pituitary-gonadal axis

Basal LH and FSH serum levels were higher than normal before E2 treatment and decreased during both phase 2 and phase 3 (Table 1). The lowest values of LH and FSH were achieved during phase 2 when the highest dose of E2 was used (25 μg twice weekly). In this phase, LH values were within the normal range, but FSH values, although reduced, remained above the normal range. GnRH administration induced a normal LH and FSH response in all phases of the study and the response was inversely related to the dose of transdermal E2 administered, the peak and the incremental area under the curve after GnRH administration being lowest in phase 2 (25 μg twice weekly) (Fig. 1 and Table 1).

The ratios between bioactivity and immunoreactivity of both circulating LH (LH B/I) and FSH (FSH B/I) were unchanged during the study (Table 1).

Basal serum levels of α-subunit did not differ among the three phases of the study protocol (Table 1). The secretory pattern of α-subunit after GnRH infusion was similar to that of gonadotropins in the three phases of the study (Table 1).

Serum T concentrations were inversely related to serum E2 levels, showing the highest levels before treatment and the lowest levels during the maximal E2 dosage (Table 1). Basal inhibin B levels were higher before transdermal E2 treatment (phase 1) and slowly decreased in phase 2 and phase 3 (Table 1). No difference in SHBG levels was found during the study (Table 1).

GH/IGF-I axis

The patient showed impaired GH responses to provocative tests consistent with a condition of GH deficiency (21) because the GH peak was less than 3 μg/liter after ITT and less than 9 μg/liter after GHRH plus ARG test (Fig. 2). Estrogen treatment at any dose did not significantly modify GH response to provocative tests (Fig. 2 and Table 2), the peak and the incremental area under the curve after provocative tests being not significantly changed during the whole protocol study. No significant change in IGF-I levels that were in the lower limit of the normal range was detected throughout the whole study (Table 2).

Pituitary-thyroid axis

Basal TSH, FT₃, and FT₄ were within the normal range and did not differ significantly among the three phases. Estrogen treatment did not modify TSH and α-subunit response to TRH administration (Table 2).

Pituitary-PRL axis

Values of basal serum PRL were at the lowest limit of the normal range during the whole study. Serum PRL levels rose after TRH administration without significant differences among the three phases of the study protocol (Table 2).

Pituitary-adrenal axis

No significant differences in basal and stimulated plasma ACTH and cortisol levels after both ITT and CRF test among the three phases of the study protocol were observed. During the three phases, basal ACTH levels were 3.1, 3.6, and 2.9 pmol/liter in phases 1, 2, and 3, respectively, and cortisol levels were 424, 610, and 502 nmol/liter, respectively. ACTH peaks after ITT were 20.9, 18.3, and 27.9 pmol/liter in phases 1, 2, and 3, respectively, and ACTH peaks after CRF were 11.3, 8.6, and 9.5 pmol/liter, respectively. Cortisol peaks after ITT were 905, 949, and 1214 nmol/liter during the three phases, and cortisol peaks after CRF were 878, 508, and 728 nmol/liter, respectively.

Discussion

The purpose of this study was to establish the role of estrogens on pituitary function in an adult male with aromatase deficiency, a unique natural model in which the effects of both congenital estrogen deprivation and estrogen replacement therapy can be studied. In this work, estrogen replacement therapy was performed using amounts of estrogens that ensured serum physiological E2 levels for an adult male. The present study clearly demonstrates that estrogens exert a complex action on pituitary function in the human male.

First, estrogens influenced not only the basal secretion of gonadotropins but also their responsiveness to GnRH. In humans evidence for the role of estrogens in modulating gonadotropin feedback mechanism has been provided from studies carried out in men with congenital estrogen deficiency. Accordingly, results from men with estrogen resistance or aromatase deficiency (1–4) confirm previous data obtained in healthy men using aromatase inhibitors or nonaromatizable androgen (22, 23, 25). Despite the presence of normal or elevated serum T, in these subjects serum gonadotropins were found elevated or in the upper limit of the normal range (1–3), indicating that aromatization of T to E2 is required for a normal functioning of gonadotropin feedback. However, in these studies, the role of estrogens on gonadotropin feedback was inferred only on the basis of hormonal levels in basal conditions. In the present study, we show that in our patient with aromatase deficiency the administration of estrogens not only reduced basal gonadotropin levels but also reduced LH, FSH, and α-subunit secretion in response to GnRH, both effects being dose dependent. However, a complete normalization of serum FSH during E2 treatment was not achieved in the presence of physiological levels of circulating E2 and supraphysiological levels of estrogens were necessary to obtain FSH normalization (3, 5). This result is probably owing to the concomitant severe impairment of spermatogenesis in our patient, a situation that explains also the finding of serum inhibin B below or at the lower limit of the normal range (26).

The observation that estrogen decreased gonadotropin responsiveness to GnRH in this patient is consistent with a recent study showing that in normal men estrogen depletion by aromatase inhibitors increased LH pulse amplitude (9, 27). Accordingly, the impact of estrogen action on gonadotropin secretion at the pituitary level has been recently demonstrated to operate from early and midpuberty to old age in men (28, 29). We did not find any change in the ratio between bioactivity and immunoreactivity of both FSH and LH in the different phases of the study, demonstrating that estrogen replacement therapy did not modify B/I ratios. Of course, our findings are restricted to the peculiar natural model of congenital aromatase deficiency. Using the rat interstitial T bioassay, Veldhuis and Dufau (30) found that in healthy young men, the E2 infusion under steady-state condition led to a decrease of the mean serum B/I LH ratio. An even more evident decrease in the mean serum B/I LH ratio was demonstrated in a man with hyperestrogenemia and very low T concentrations associated with an adrenal tumor (31). Taken together, these findings provide support for the idea that endogenous estrogen feedback action results in decreased secretion of biologically active LH in man (32). Further studies should help to clarify this issue.

In this patient the expected enhancement of GH secretion secondary to estrogen administration (12) did not occur. Indeed, the patient showed a severe impairment of GH secretory response to potent provocative stimuli, thus suggesting an alteration of the mechanisms involved in GH secretion at both pituitary and hypothalamic level. Even though the secretion of GH is impaired, no signs or symptoms of GH deficiency were seen. In fact, the patient was tall (189 cm), and serum IGF-I concentrations were in the normal range. Therefore, it appears that GH secretory dysfunction was probably not present during childhood and during puberty, but it is an adult-onset phenomenon. Nevertheless, these results allow us to speculate on the relationships between estrogens and the mechanisms controlling GH secretion.

Sex steroids enhance pituitary GH secretion during male puberty, an event that is secondary to androgen aromatization to estrogens (33). Recently, in fact, it has been demonstrated that the ER-α gene is expressed in GHRH neurons suggesting that locally produced estrogens may modulate GHRH secretion in the rat hypothalamus (34). In addition, α-estrogen receptor knockout, αβ-estrogen receptor knockout, and aromatase knockout male mice show impaired GH/IGF-I axis because of the congenital lack of estrogen action (6, 35). At present we ignore what is the real impact of estrogens on maturation of pituitary GH/IGF-I axis early during male development and/or at the time of puberty. However, some clinical evidences speak in favor of a strict linkage between estrogen and GH/IGF-I axis maturation at least at puberty (33). In this view it is possible to speculate that in our patient congenital severe estrogen depletion may have contributed to a partial disruption of somatotroph development, eventually resulting in GH deficiency and insensitivity of GH/IGF-I axis to circulating estrogens in adult life. Even though the secretory reserve of GH seems to be impaired, daily GH secretion was probably able to ensure normal IGF-I levels. Nevertheless, serum IGF-I levels were not modified by estrogen treatment and remained at the low limit of the normal range. However, this finding does not agree with previous data showing a positive modulation of estrogen on IGF-I serum levels in both animals (35) and humans (10, 13, 33) coupled with a rapid activation of the IGF-I receptor pathway through ERα (36). Finally, these data suggest that, in clinical practice, caution is needed in evaluating the results of provocative pituitary tests for the evaluation of GH pituitary secretion in male subjects with congenital estrogen deficiency.

Both TRH-stimulated PRL and TSH levels were not modified by estrogen supplementation, suggesting that supraphysiological (16, 17) but not physiological doses of estrogens are required to enhance PRL and TSH pituitary secretion after TRH injection. However, PRL serum levels in basal conditions were below or at the lower level of the normal range for normal men during the study. Similarly, the TRH-stimulated TSH levels were slightly reduced if compared with those recorded in normal men (17). Data available from animal models of estrogen deficiency suggest that estrogens modulate both PRL gene expression in the anterior pituitary (37, 38) and lactotroph cell growth (38). A wide degree of impaired GH, PRL, and TSH pituitary secretion in a man with a congenital lack of estrogen action suggest some implications on the role of estrogens on pituitary cells, which probably belong to the same cell lineage from a unique progenitor (39, 40). Further studies are needed also to provide data on the relationship between estrogens and Pit-1 gene expression (37, 41), which could account for the regulation of the development of GH-, PRL-, and TSH-producing cells as well as their function during lifetime (39, 40, 42). Interestingly, estrogen replacement therapy did not restore the impaired pituitary secretion of GH, PRL, and TSH in our patient, in accordance with a putative role of estrogens in the modulation of pituitary cell function early during the maturation of hypothalamic-pituitary unit.

In conclusion, aromatization of T to E2 is required at pituitary and hypothalamic level for a normal functioning of gonadotropin feedback in both basal and stimulated conditions. Furthermore, the congenital lack of estrogen activity seems to be associated with an impaired secretion of GH, and to a lesser extent of PRL and TSH, thus suggesting a possible role of estrogens on the pituitary secretion of these hormones in the human male.

Abbreviations:

 
• ARG,

  Arginine;

 
• B/I,

  bioactivity/immunoreactivity ratio;

 
• CRF,

  corticotropin-releasing factor;

 
• FT3,

  free T₃;

 
• FT4,

  free T₄;

 
• ITT,

  insulin tolerance test.

References