Ovarian activity before and after gonadal suppression by GnRH-a in patients with polycystic ovary syndrome , hyperandrogenism , hyperinsulinism and acanthosis nigricans

OBJECTIVE. To investigate the ovarian activity before and after gonadal suppression with GnRH-analog in patients with PCO, hyperandrogenism, hyperinsulinism and acanthosis nigricans. Design: Controlled clinical study. Setting: Tertiary academic medical center. Patients: Six patients with clinical findings of PCO, hirsutism and acanthosis nigricans. INTERVENTIONS. Morning blood samples in the follicular phase to determine the steroid levels, glucose and insulin curve, comparing to a control group. Administration for 2 consecutive months of a GnRH-analog, comparing, in the study group, the free testosterone levels before and after ovarian suppression. MAIN OUTCOME MEASURE. Determination of insulin levels in PCO, hirsutism and acanthotic patients and the free-testosterone levels before and after gonadal suppression. RESULTS. Insulin levels were significantly higher in the study group when compared to normal women during the glycemic test. We also found a significant decrease in the free-testosterone levels after 2 months of gonadal suppression with GnRH-analog when compared to the initial time. CONCLUSIONS. Patients with PCO, hirsutism and acanthosis nigricans present high levels of insulin, suggesting an ovarian hyperesponsiveness, which is not sustained when gonadotrophic blockage was achieved.


INTRODUCTION
Sclerotic modifications of the human ovary have been described for more than 100 years 1 , and the association between menstrual alterations, hyperandrogenism (HA) and infertility was acknowledged by Stein and Leventhal in 1935 2 .Since that first report, many studies have followed and despite gains in the knowledge, the pathophisiology of the polycystic ovary syndrome (PCOS) remains unclear.
Ovarian steroidogenesis is regulated chiefly by the hypophyseal gonadotrophins.However, it is also influenced by a number of peptides, so-called growth factors, which modulate follicular growth [3][4][5][6] .Among these, insulin and the insulin-like growth factor-1 (IGF-1) should be highlighted 7 .The presence of raised serum insulin levels in patients with PCOS and hyperandrogenism was initially elucidated by Burghen et al. 8 and Taylor et al. 9 .Nevertheless it was Barbieri and Ryan 10 , who, in 1983, described a group of patients with PCOS, HA, peripheral insulin resistance (IR) and acanthosis nigricans, naming it as HAIR-AN syndrome.According to the authors, the insulin resistance raised circulating insulin levels, which, in turn, excess would act in the ovaries and epidermis.Thus, there would be an interaction between insulin receptors and the receptors of IGF-1 in the ovary and skin, leading to androgen production and epidermal proliferation, resulting in hirsutism and acanthosis nigricans respectively.Since these early descriptions, many reports of a correlation between hyperinsulinism and hyperandrogenism in PCOS 11 have followed.
We hypothesized that ovarian hyperandrogenism was critically dependent upon pulsatile gonadotropin release and would not be sustained by hyperinsulinemia alone.So, to test this hypothesis, we compare the response of women with PCOS after two months of GnRH-analog (GnRH-a) to eumenorrheic women.

METHODS
Twelve women were studied at the Endocrinologic Gynecology Section of the Discipline of Gynecology of Escola Paulista de Medicina, São Paulo, Brazil.These were classified into two study groups, I and II.The first was composed by six eumenorrheic women, without previous gynecological and/or endocrine alterations.The second study group consisted of another six women with polycystic ovary syndrome, hyperandrogenism, peripheral resistance to insulin and acanthosis nigricans.All women were between the age of 15 and 35 years, have not been taking hormonal medications in the previous six months.
All underwent thorough gynecological examination.The clinical history of those patients in Group II revealed primary oligomenorrhea, hirsutism and infertility.The diagnosis of acanthosis nigricans was based upon the presence of raised hyperpigmented and papillary lesions in the axillary region, nape of the neck, as inframammary sulcus.All were obese with body mass index (BMI) over 25 12 .Clinical features are displayed in table.
In all cases the following procedural routine was adapted.

Laboratory study
Biochemical profile whereby hemogram, proteinogram, urea and creatinine were determined since there are raised insulin levels in some renal diseases 11 .The hormonal profile consisted in the dosage of levels of prolactin (PRL), growth hormone (GH), cortisol (C), thyroid-stimulating hormone (TSH), and thyroxine (T4) assessed by the conventional radioimmunoassay (RIA) technique.
All of these tests were carried out by immunofluorimetric technique, except for free T, which was done by RIA.All of the hormone level assessments were carried out in duplicate and the variation coefficients were less than 10% intraassay and less than 12% inter-assay.Table shows the hormonal profile of the patients in the Study Group.

Oral glucose tolerance test
The oral glucose tolerance text (OGTT) was carried out to determine glycemic and insulinemic curves.The patients in both groups were told to ingest a diet rich in carbohydrates (more than 150g) for 3 days prior to the test 13 .After an overnight fast of 10-12h, basal blood sampling was started, followed by administration of 75g of glucose solution.Samples were taken at times 0, 15, 30, 60, 90 and 120 minutes.

Control of ovarian suppression by analogue agonist of the gonadotropin releasing hormone (GnRH-a)
The patients in Group II received two deep intramuscular injections of GnRH-a (D-Trp6-GnRH) for gonadal suppression at 30 day intervals.Assays were carried out 30 days after each injection to measure levels of free T, E 2 , FSH, and LH 14 .
With the exception of the assays carried out after the control of ovarian suppression, all the others, including OGTT were carried out during early follicular phase (2nd to 5th day), or randomly in the amenorrheic patients.
Data were analyzed statistically by nonparametric tests by the test of Friedman 15 rank analyses of variance and the Mann-Whitney test for two independent samples 16 .Significance levels was considered at P < 0.05.

RESULTS
Figure 1 shows the average of glucose and insulin results respectively for each group during OGTT.On analyzing each group separately, one sees that the glycemic levels rose rapidly after 15 minutes, i.e., soon after caloric intake.These levels are significantly higher than the ones observed at 90 and 120 minutes (P<0.05).In Group II, however, despite the same sudden initial rise after 15 minutes, the glycemic values stayed high at 90 and 120 minutes.The comparison of glycemic behavior between the two groups, made by the Mann-Whitney test, revealed to be statistically higher at 90 minutes in Group II.
The second part of the figure shows the insulin determinations in the two groups during the same test.When each group is examined individually one notes that in Group I, the insulin variation took place later, i.e., after 30 and 60 minutes (P<0.05).These levels are considerably higher than the initial ones.As for Group II, the insulin levels rose significantly after 30, 60 and 90 minutes (P<0.05).The Mann-Whitney test shows that all levels of insulin in Group II are statistically higher, at the different times, if compared to those of the control group.
Figure 2 shows simultaneous curves for glucose and insulin in both groups.In Group I the average of insulin levels corresponds to one third of those of glucose although in Group II they are similars.In Figure 3 the serum levels of testosterone and estradiol prior to and after one and two months following administration of GnRH-a.One can note that, in comparison to the basal levels, there has been a marked fall in the respective levels.

DISCUSSION
It has become evident, since the 80's, that ovarian function is not regulated only by the hypophyseal gonadotrophins, but is also influenced by action of other hormones and growth factors 17 .Among these, insulin and IGF-1 have been the most extensively studied.
The first data about this association have been known since 1921 with the so-called Archard-Thiers Syndrome, or bearded diabetes, where raised serum insulin levels and hyperandrogenism have been associated.
There have also been frequent reports associating conditions of marked insulin resistance, and therefore hyperinsulinism, with exacerbated gonadal activity, such as premature puberty and hirsutism.Conversely, insulin-dependent diabetic patients, i.e., hypoinsulinemic, frequently present late menarche and sexual infantilism 7 .
The association between hyperinsulinism and PCOS was clarified by Burghen et al. 8 These researchers suggested that raised testosterone levels were a consequence of the direct stimulatory effect of insulin on the female gonad.Barbieri and Ryan 10 , when describing the HAIR-AN Syndrome, highlighted a "nature's experiment", in which these patients were characterized by chronically raised insulin levels, thus denoting ovarian and epidermic hyperactivity.
It is known that acanthosis nigricans is related to various hyperinsulinemic conditions, such as non-insulin-dependent diabetes, Addison's disease, obesity and PCOS 17 .The factors which promote epidermal proliferation have not been completely elucidated, but the main hypothesis is based upon stimulation through growth factors such as melanocyte-stimulant hormone, growth hormone, adrenocorticotrophic hormone and insulin, among others 10 .In the human ovary, current evidence, both in vitro and in vivo allow us to state that insulin connects to specific receptors or to IGF-1   receptors, stimulating theca and stromal cells to increase steroidogenesis 18 .
Thus in this study we set out to evaluate gonadal function in patients with PCOS, hyperandrogenism and insulin resistance.Initially a group of patients with these characteristics and who were obese were selected.Among the various clinical conditions related to peripheral resistance to insulin, obesity seems to be the most prevalent and widely studied 19,20 .
Between the responsible mechanisms it is accept the decrease number of receptors (downregulation) and alteration of post-receptor activation 21 .To date, in spite of populational studies which relate insulin levels to those of glucose [22][23][24][25] , it should be noted that there are no criteria which safely establish hyperinsulinism 26 .Thus we have attempted to compare the study group to the control group, which consisted of eumenorrheic, non-hirsute and non-obese females with BMI <25 (Group I).During OGTT it was noted that, in Group I, after the stimulus of ingest of carbohydrates there is rapid rise in glycemia, noticeable at 15 minutes, followed by a gradual drop towards the end of the test.These data are concordant with wide populational studies 13 .This also suggests that after stimulation there would be insulin liberation, which would promote glucose mobilization.
In the patients with PCOS (Group II), one initially observes similar behavior.However, at the end of the test their levels are discretely higher in comparison to the time of 15 minutes, even comparing a carbohydrate-intolerance state 13 .These data in themselves suggest that the hypoglycemiant hormone be unable to promote the entry of glucose into the cell.
As for the insulin curve, in Group I, its levels are lower than their respective glycemic levels, i.e., about 1/3, rising significantly at 30 and 60 minutes in relation to the initial time.In Group II the basal levels are also lower than those of glucose, but after the stimulus there is an intense rise, reaching its peak at 30 minutes, when it surpasses the corresponding glycemic value.This characteristic suggests the hyperinsulinic state 24,27 .When the insulin levels in the two groups are compared, there is statistical difference in all the times studied, thus characterizing hyperinsulinism.
In reference to the use of GnRH-a, these drugs constitute powerful agents, capable of promoting specific ovarian suppression 28 .In our material the medical castration was proven by the significant drop in estradiol levels even in the first month after administration.As for hyperandrogenism a statistically significant drop in the levels of free testosterone was observed at the end of the second month of treatment, similarly to the findings of Geffner et al. 29 .Such fact indicates the ovarian source is responsible for the increased production of androgens in patients with PCOS, hyperandrogenism, hyperinsulinism and acanthosis nigricans.
Therefore, the present study enables us to state that the serum insulin levels in normal women correspond to approximately one third of the glycemic levels.However, in the patients with PCOS, hyperandrogenism, hyperinsulinism and acanthosis nigricans the average insulin levels were always high, nearing the average glucose levels.Furthermore, in these patients, after gonadal blockage there is a drop in androgen levels, which allow us to state that the ovaries represent, most likely, the source of these sexual steroids.

Fig. 1 -
Fig. 1 -Serum glucose and insulin during the OGTT.Levels of glucose are greater in group II comparing to Group I (P<0.05) at 90 minutes.Levels of insulin are greater in Group II (P<0.05) in comparison to Group II at all times.

Fig. 2 -
Fig. 2 -Average levels of glucose and insulin in each group duringOGTT.

Fig. 3 -
Fig.3-Levels of estradiol and free-testosterone before and after 1 and 2 months with GnRH-a in the patients of Group II.After 2 months both levels are statistically lower (P<0.05).