madman
Super Moderator
Impact of Clomiphene Citrate on the Steroid Profile in Dysmetabolic Men with Low Testosterone Levels (2021)
Carla Pelusi, Flamina Fanelli, Margherita Baccini, Giovanni De Pergola, Vincenzo Triggiani, Marco Mezzullo, Alessia Fazzini, Guido Di Dalmazi, Biljana Petrovic, Paola Paterini, Antonio Maria Morselli Labate, Uberto Pagotto, Vito Angelo Giagulli
Abstract
Clomiphene citrate (CC) in male hypogonadism increases testosterone (T) and estrogen levels by stimulating pituitary gonadotropin release. Our group confirmed these hormonal changes in a randomized, cross-over, double-blind trial of CC versus placebo in addition to metformin, conducted in 21 obese dysmetabolic men with low T levels. However, we hypothesize that based on its mechanism of action, CC may directly or indirectly affect adrenal steroidogenesis. The aim of this sub-study was to better understand the changes in steroid levels and metabolism induced by CC treatment. We assessed 17α-hydroxypregnelone (17αOH-P5), dehydroepiandrosterone (DHEA), progesterone (P4), 17α-hydroxyprogesterone (17αOH-P4), androstenedione (A), T, dihydrotestosterone (DHT), estrone (E1), 17β-estradiol (E2), 11-deoxycortisol (11S), cortisol (F), and cortisone (E) by LC-MS/MS, and corticosteroid-binding globulin (CBG) by ELISA, before and after each treatment. In addition, free-F and steroid product/precursor ratios were calculated. We observed a significant change in serum levels induced by CC compared with placebo for 17αOH-P4, DHT, T, E2, E1, F, E, and CBG, but not free-F. In addition, compared to placebo, CC induced higher 17αOH-P4/P4, E2/E1, 17αOH-P4/17αOH-P5, A/17αOH-P4, T/A, E1/A, F/11S, and F/E ratios. Therefore, besides the CC stimulating effect on testis steroidogenesis, our study showed increased F, E, but not free-F, levels, indicating changes in steroid metabolism rather than adrenal secretion stimulation. The steroid profiling also revealed the CC stimulation of the Δ5 rather than the Δ4 pathway, thus indicating considerable testicular involvement in the increased androgen secretion.
Introduction
Selective estrogen receptor modulators (SERMs) have attracted great interest in restoring endogenous testosterone (T) in adult males suffering from functional hypogonadotropic hypogonadism [1–4]. Our group [5] and others [6] have shown that clomiphene citrate (CC) reactivates the hypothalamic-pituitary-gonadal (HPG) axis, thereby promoting an increase in endogenous T levels in obese men with low T levels with and without dysmetabolic abnormalities.
In a previous work [5], we demonstrated that CC therapy-induced LH and FSH release and stimulated the testis, thereby determining an increase in serum T, dihydrotestosterone (DHT), and 17β-estradiol (E2) levels. However, although the T level is restored to physiological concentration in these subjects during CC treatment, other hormones increased compared to normal ranges with a downstream effect. In fact, few studies have shown an increase in cortisol (F) levels during CC therapy, thus also postulating an activation in the adrenal steroidogenesis [7–10]. However, these studies are old, dating back to the 1970s, with different assays, and involving very few participants.
The mechanism underlying the possible adrenal activation could be explained by the increased LH and/or estrogen levels induced by CC treatment. In fact, they may act directly, by binding specific receptors expressed at the adrenal gland level [11–14], or indirectly, through the estrogen stimulation of the hepatic production of corticosteroid-binding globulin (CBG) [15].
Several studies have shown that high levels of LH exert subtle changes in adrenal function in postmenopausal women thus promoting cortisol secretion [16,17]. It has also been suggested that the activation of the estrogen receptor α (ERα) could enhance cortisol production and adrenocortical tumorigenesis, which is probably involved in the development of Cushing syndrome during pregnancy [18]. In addition, pregnancy and the estrogen-based combined oral contraceptive pills increase CBG concentrations two to three-fold, resulting in sustained physiologic hypercortisolism [19].
*Based on these assumptions, we decided to extend our study on the effect of CC treatment on the adrenal cortex by investigating the biosynthesis and metabolism of glucocorticoid and androgens. We thus investigated the modifications in serum levels of corticosteroids, CBG, and other sex hormones besides T. In addition, we hypothesized that the evaluation of these steroidogenesic pathways, carried on by the enzyme sets belonging to both adrenal and gonadal glands [20, 21], might help in highlighting which gland is the main CC target. To explore the synthetic and metabolic steroid processes impacted by the CC treatment, we analyzed the modifications of the product-over-precursor ratios in the androgen and glucocorticoid pathways as the possible indirect estimates of enzyme activity.
Steroid measurements
The serum steroid profile was assayed by two liquid chromatography-tandem mass spectrometry (LC-MS/MS) methods. The first method included cortisone (E), F, 11-deoxycortisol (11 S), 17αOH-progesterone (17αOH-P4), progesterone (P4), dehydroepiandrosterone (DHEA), androstenedione (A), and T [23], while the second included estrone (E1), E2, DHT, and 17αOH-pregnenolone (17αOH-P5) [24]. Four serum samples were analyzed for each of the 21 eligible subjects, and all samples were measured in two consecutive runs. Due to analytical specificity requirements, as previously detailed [23, 24], data were discarded relating to 17αOH-P5 in five samples from four subjects, to P4 in three samples from one subject of E2 in three samples from three subjects, and to DHT in one sample. In addition, a very small serum volume was available for one sample, resulting in the sensitive measurement of only the T value.
We measured plasma CBG using the Biovendor ELISA kit (RD192234200R; Brno, Czech Republic). Free cortisol (free-F) was measured by the law of mass action according to the formula free-F (nmol/l) = 1000 * ((Z2 + 0.0122 C)0.5–Z), where Z = 0.0167 + 0.182 × (F-CBG), with F and CBG in μmol/l [25, 26].
Steroid enzymatic pathways
We analyzed the following steroid product-over-precursor molar ratios as indirect surrogate measures of the respective enzymatic activities: a) A/17αOH-P4 and DHEA/17αOH-P5 ratios for 17α-hydroxylase/17,20-lyase (CYP17A1) in Δ4 or Δ5 pathway, respectively; b) 17αOH-P4/P4 ratio for 17α-hydroxylase (CYP17A1); c) A/ DHEA and 17αOH-P4/17αOH-P5 ratios for 3β-hydroxysteroid dehydrogenase type 2 (HSD3B2); d) T/A and E2/E1 ratios for 17β-hydroxysteroid dehydrogenase type 3 (HSD17B3) and type 1 (HSD17B1), respectively; e) DHT/T ratio for 5α-reductase (SRD5A); f) E1/A and E2/T ratios for aromatase (CYP19A1); g) 11S/17αOH-P4 ratio for 21-hydroxylase (CYP21A2); h) F/11 S ratio for 11β-hydroxylase (CYP11B1), and i) F/E for 11β-hydroxysteroid dehydrogenase type 2 (11βHDS2) [20].
Steroid levels
Pre-treatment steroid profile
There were no significant differences in basal estrogen levels (E1 and E2), glucocorticoids (F, E, and 11 S), androgens (T, DHT, DHEA, and A), 17αOH-P5, CBG, and free-F in the basal samples collected before CC and Plac treatments (▶Tables 1 and ▶2). This highlighted the good efficacy of the drug wash-out period for these hormones. However, the serum 17αOH-P4 level before CC was significantly higher than before the Plac treatment (▶Table 1).
Effects of the CC and Plac treatments on steroid levels
▶Table 3 shows the fold changes in steroid levels induced by CC and Plac at the end of the 3-month treatment. The CC therapy increased serum P4 (p=0.044), 17αOH-P4 (p<0.001), A (p=0.025), T (p < 0.001), DHT (p < 0.001), E1 (p < 0.001), E2 (p < 0.001), F (p<0.001), and E (p<0.001), while the Plac treatment did not significantly change the serum corticosteroids, estrogens and androgens, except for an increase in 17αOH-P5 (p=0.039) and T (p=0.030) levels. In addition, CBG (p< 0.001), but not free-F, increased significantly during the CC treatment (▶Table 2). When the fold changes induced by CC were compared with those observed after placebo, we found that the effect of CC treatment on T, E2 (both p < 0.001), 17αOH-P4, DHT (both p = 0.001), E1, F (both p = 0.002) and E (p = 0.028) was significantly higher.
Steroid product-over-precursor ratios
Profile of pre-treatment product-over-precursor ratios
Steroid product-over-precursor ratios observed before the CC treatment were not significantly different from those observed before the Plac treatment, except for the CYP17A1 activity (A/17αOHP4), which was significantly lower before CC than before the Plac (p = 0.023).
Effects of the CC and Plac treatments on steroid productover-precursor ratios
▶Table 4 shows the effects of both CC and Placebo treatments on the steroid product-over-precursor ratios. CC significantly increased CYP17A1 (17αOH-P4/P4, A/17αOH-P4), HSD17B3(T/A), HSD17B1(E2/E1) (all p < 0.001), HSD3B2 (17αOH-P4/17αOH-P5), 11βHDS2 (F/E) ,(both p=0.001), HSD3B2 (17αOH-P4/17αOH-P5, p=0.001; A/DHEA p=0.005), CYP19A1 (E1/A, p=0.006), CYP21A2 (11S/17αOH-P4, p=0.002) and CYP11B1 (F/11S, p<0.001) activities, whereas the Plac did not significantly affect any of the analyzed ratios. In particular, the effect of CC treatment on CYP17A1 (17αOHP4/P4, p<0.001; E2/E1, p=0.001), HSD3B2 (17αOH-P4/17αOH-P5, p = 0.009), HSD17B3 (A/17αOH-P4, p = 0.004), HSD17B1 (T/ AP=0.003), CYP19A1 (E1/A, p=0.021), CYP11B1 (F/11S, p=0.008), and 11βHDS2 (F/E, p=0.023) was significantly higher than the effect of Plac.
Discussion
It is generally accepted that CC, a nonsteroidal estrogen agonist/ antagonist belonging to the SERM family, inhibits hypothalamic estrogen feedback, thereby increasing gonadotropin-releasing hormone (GnRH) production and stimulating gonadotropin secretion, which, in turn, results in increasing serum T levels by stimulating the Leydig cells [27]. To date, a larger number of studies have confirmed this CC effect on testicular steroidogenesis [28]. In fact, CC improves serum levels of LH and T in patients affected by a functional deficit of T, as it does in dysmetabolic conditions (i. e., T2DM and/or obesity) [5,6] with and without infertility [29], also in young or elderly men [30] and those affected by erectile dysfunction [31], in the absence of severe organic testicular damage [2].
CC could therefore be an alternative and probably more appropriate treatment than testosterone replacement therapy in patients affected by functional hypogonadism and high cardiovascular risk such as dysmetabolic men [32, 33].
Notably, evidence on the effects of CC on the overall testicular steroidogenesis is well established, yet the evidence on its effect on adrenal steroidogenesis is scant even though it has been demonstrated that both glands express the LH and estrogen receptors and could be affected by CC treatment [34–36]. To examine the enzymatic pathways behind the CC therapeutic effect, we thus extended our steroid analysis to a broad serum profile, as facilitated by the state-of-the-art LC-MS/MS technique, in the study cohort previously investigated in the CC therapy-controlled study [5]. We also measured CBG concentrations to assess the potential CC effect on both serum total F and its free fraction, which is known as the active hormonal form and the reliable expression of adrenocortical function [37].
We first ascertained that patients exhibited similar baseline circulating steroid levels before undergoing each treatment. Only 17αOH-P4 was higher before the CC treatment in comparison with the Plac phase. However, we found no explanation for such an isolated hormone increase. To explore the possible reason for this observation, we analyzed the carry-over effect between the two interventions in the two arms and found no significant impact (data not shown). Our data thus led us to attribute the difference in 17αOH-P4 basal levels to chance.
*In our previous studies, we showed that, besides T, CC therapy results in increasing serum DHT, E2, the sex hormones binding globulin (SHBG), LH, and FSH [5,38]. In the present study, we added the novel finding that CC administration along with Met significantly increases circulating E1, 17αOH-P4, F, E, and CBG, but not other steroids including androgen and glucocorticoid precursors, such as 17αOH-P5, A, DHEA, and 11 S levels (see ▶Table 3).
Notably, the sole administration of Met had no impact on the hormone levels examined. The CC-induced CBG increase was expected and potentially explained both by the estrogen stimulation of hepatic protein production [15] and by CC direct estrogen-like activity [39]. Moreover, the increase in CBG levels (approximately 30%) did not exceed the normal range (300–700 nmol/l) [40] and the calculated free-F levels were unchanged. These data, in addition to the unchanged adrenal androgen (DHEA, A, and 17αOHP5) and corticosteroid (11S) blood levels observed during CC treatment, suggested a lack of any direct adrenal stimulation by this drug [8]
To understand which pathway was affected the most by the CC treatment, we analyzed the molar ratios derived from the product-over-precursor steroid pairs as an indirect estimate of the corresponding enzyme activity in androgen and glucocorticoid synthesis, as well as in their metabolism in target tissues. In our cohort, CC therapy combined with Met significantly increased A/DHEA and T/A ratios, while reducing the A/17αOH-P4 ratio but not the DHEA/17αOH-P5 (see ▶Table 4). These data suggest a possible increase in the enzyme activity of 3β-hydroxysteroid dehydrogenase type 2 and 17β-hydroxysteroid dehydrogenase type 3 and an unclear suppression of the 17α-hydroxylase/17,20-lyase activity. The possible steroid pathway CC-induced is summarized at ▶Fig. 1.
As the adrenal glands are rich in the 17,20-lyase but poor in the 17β-hydroxysteroid dehydrogenase type 3 enzyme [20, 41], our data point to the CC effect at the testicular level, thereby indicating that the main effect is on the Δ5 pathway, rather than the Δ4 pathway for androgen synthesis [42]. Interestingly, our data indicate that CC therapy seems to act on testis steroidogenesis in a similar way to the hypothesis put forward for human chorionic gonadotropin (hCG). In fact, hCG stimulation in healthy and fertile males (doses ranging from 1500 UI to 5000 UI im) appears to enhance T secretion above all by facilitating the Δ5 pathway, due to the reduction in CYP17 activity (17α-hydroxylase/17,20-lyase) (Δ4) [43, 44].
In our patients, the CC therapy succeeded in increasing estrogens (E1 and E2), DHT, and the CYP19A1 (E1/A) activities, whereas the SRD5A (DHT/T) remained unchanged, thus suggesting that aromatase and 5α-reductase are only partially stimulated or unaffected by CC, respectively.
Our results showing the stimulating effects of CC therapy on aromatase activity agree with previous data obtained in oophorectomized women [9] and normal men [8]. In addition, we showed that CC elicited a higher than two-fold increase in serum E2. This could be due to the reduced hormone excretion, given a concomitant increase in serum SHBG by CC in our patients [5]. However, the affinity of E2 for SHBG is much lower than in the active androgens T and DHT (30% and 50%, respectively) [45], which are also increased by CC treatment, thus suggesting that the reduced excretion plays a limited role in the serum total E2 increase. On the other hand, the peripheral conversion of T into E2 is known to explain up to 40% of E2 levels [46], which would suggest only a limited impact of peripheral aromatase on the E2 level observed upon CC treatment. Therefore, the main source of the CC-induced increase in total serum E2 in our patients may, in fact, be the aromatase activity at the Leydig cell level [47, 48]. A reduced ratio between circulating T and E2 levels has been associated with damage in both the Leydig and Sertoli cells (primary hypogonadism) as well as with the failure of aromatase inhibitors or anti-estrogens, such as CC, to improve semen parameters and T levels [49,50]. We found that CC treatment induced a parallel increase in circulating T and E2 without altering their balance. In fact, these results highlight that there was no impairment in testicular steroidogenesis in our patients.
The increase in DHT levels that we observed after CC therapy does not appear to involve an increased 5α-reductase activity, as the DHT/T ratio was unchanged. In addition, the conversion rate of T into DHT by 5α-reductase has been reported to be 3.5 % [51]. Given such a conversion rate and considering the serum T levels obtained after the CC therapy (18.7 nmol/l), the estimated amount of DHT that should derive from T conversion would be around 0.7 nmol/l, which is twice the increase in serum DHT that we found at the end of the CC treatment.
These data also enable us to exclude the activation of the peripheral backdoor pathway to DHT, although the rise in serum 17OHP4 (40 %) could facilitate its peripheral metabolism [20]. Therefore, the increase in DHT level that we observed after CC therapy is probably not the result of an increase in its synthesis but is rather due to the reduction in its metabolic clearance, as supported by the previously mentioned increase in SHBG levels [5]
Finally, CC treatment increased the serum levels of 17αOH-P4, E, and F by 1.2–1.6 fold. In particular, 3-month CC therapy increased serum F levels (425 ± 113 nmol/l) compared to the basal levels (275±105 nmol/l), but these F levels are still below the upper reference limit of the same assay (550 nmol/l) [22]. This was explained by the increase in CBG as shown by the unchanged calculated values of the free-F during CC treatment. A similar mechanism would also explain the increased values of E and 17OH-P4. These data are in line with a few studies on the CC effect on CBG which showed an increase in the protein related to the treatment through an estrogen-like effect [10,38]. However, we cannot completely rule out a synthesis by the adrenal gland, as apparently suggested by the CYP11B1 expression (F/11S), nor a reduction in the oxidative metabolism at the target tissues, as suggested by the reduced 11βHDS2 activity (F/E). In addition, the CYP21A2 (11S/17αOH-P4) was reduced, thus suggesting a possible compensatory protective mechanism against glucocorticoid excess induced by CC. On the other hand, it is very unlikely that there was a CC stimulation effect mediated by the increase in LH and E2 levels and the action on their respective receptors expressed at this site [36].
*To address these issues, determining the ACTH before and after each treatment phase would have been decisive, but unfortunately was not conducted and is thus the main limitation of our study.
In conclusion, in previous work, we showed that CC treatment improves the function of the HPG axis in obese dysmetabolic men with functional hypogonadism by increasing serum LH and, in turn, T and E2 [5]. In the present study, we found that the improvement in T secretion due to CC seems to be caused by the testicular Δ5 rather than the Δ4 pathway, without causing a significant increase in 5α-reductase activity at the target tissues. In addition, we showed that there is an enhancement of plasma F levels, but not of the free-F, indicating a lack of a direct effect of CC therapy on the adrenal gland.
Future studies are needed to establish whether these hormonal changes induced by CC remain or progress during longer treatment periods, which could counter the evident beneficial effects of CC treatment on testicular steroidogenesis.
Carla Pelusi, Flamina Fanelli, Margherita Baccini, Giovanni De Pergola, Vincenzo Triggiani, Marco Mezzullo, Alessia Fazzini, Guido Di Dalmazi, Biljana Petrovic, Paola Paterini, Antonio Maria Morselli Labate, Uberto Pagotto, Vito Angelo Giagulli
Abstract
Clomiphene citrate (CC) in male hypogonadism increases testosterone (T) and estrogen levels by stimulating pituitary gonadotropin release. Our group confirmed these hormonal changes in a randomized, cross-over, double-blind trial of CC versus placebo in addition to metformin, conducted in 21 obese dysmetabolic men with low T levels. However, we hypothesize that based on its mechanism of action, CC may directly or indirectly affect adrenal steroidogenesis. The aim of this sub-study was to better understand the changes in steroid levels and metabolism induced by CC treatment. We assessed 17α-hydroxypregnelone (17αOH-P5), dehydroepiandrosterone (DHEA), progesterone (P4), 17α-hydroxyprogesterone (17αOH-P4), androstenedione (A), T, dihydrotestosterone (DHT), estrone (E1), 17β-estradiol (E2), 11-deoxycortisol (11S), cortisol (F), and cortisone (E) by LC-MS/MS, and corticosteroid-binding globulin (CBG) by ELISA, before and after each treatment. In addition, free-F and steroid product/precursor ratios were calculated. We observed a significant change in serum levels induced by CC compared with placebo for 17αOH-P4, DHT, T, E2, E1, F, E, and CBG, but not free-F. In addition, compared to placebo, CC induced higher 17αOH-P4/P4, E2/E1, 17αOH-P4/17αOH-P5, A/17αOH-P4, T/A, E1/A, F/11S, and F/E ratios. Therefore, besides the CC stimulating effect on testis steroidogenesis, our study showed increased F, E, but not free-F, levels, indicating changes in steroid metabolism rather than adrenal secretion stimulation. The steroid profiling also revealed the CC stimulation of the Δ5 rather than the Δ4 pathway, thus indicating considerable testicular involvement in the increased androgen secretion.
Introduction
Selective estrogen receptor modulators (SERMs) have attracted great interest in restoring endogenous testosterone (T) in adult males suffering from functional hypogonadotropic hypogonadism [1–4]. Our group [5] and others [6] have shown that clomiphene citrate (CC) reactivates the hypothalamic-pituitary-gonadal (HPG) axis, thereby promoting an increase in endogenous T levels in obese men with low T levels with and without dysmetabolic abnormalities.
In a previous work [5], we demonstrated that CC therapy-induced LH and FSH release and stimulated the testis, thereby determining an increase in serum T, dihydrotestosterone (DHT), and 17β-estradiol (E2) levels. However, although the T level is restored to physiological concentration in these subjects during CC treatment, other hormones increased compared to normal ranges with a downstream effect. In fact, few studies have shown an increase in cortisol (F) levels during CC therapy, thus also postulating an activation in the adrenal steroidogenesis [7–10]. However, these studies are old, dating back to the 1970s, with different assays, and involving very few participants.
The mechanism underlying the possible adrenal activation could be explained by the increased LH and/or estrogen levels induced by CC treatment. In fact, they may act directly, by binding specific receptors expressed at the adrenal gland level [11–14], or indirectly, through the estrogen stimulation of the hepatic production of corticosteroid-binding globulin (CBG) [15].
Several studies have shown that high levels of LH exert subtle changes in adrenal function in postmenopausal women thus promoting cortisol secretion [16,17]. It has also been suggested that the activation of the estrogen receptor α (ERα) could enhance cortisol production and adrenocortical tumorigenesis, which is probably involved in the development of Cushing syndrome during pregnancy [18]. In addition, pregnancy and the estrogen-based combined oral contraceptive pills increase CBG concentrations two to three-fold, resulting in sustained physiologic hypercortisolism [19].
*Based on these assumptions, we decided to extend our study on the effect of CC treatment on the adrenal cortex by investigating the biosynthesis and metabolism of glucocorticoid and androgens. We thus investigated the modifications in serum levels of corticosteroids, CBG, and other sex hormones besides T. In addition, we hypothesized that the evaluation of these steroidogenesic pathways, carried on by the enzyme sets belonging to both adrenal and gonadal glands [20, 21], might help in highlighting which gland is the main CC target. To explore the synthetic and metabolic steroid processes impacted by the CC treatment, we analyzed the modifications of the product-over-precursor ratios in the androgen and glucocorticoid pathways as the possible indirect estimates of enzyme activity.
Steroid measurements
The serum steroid profile was assayed by two liquid chromatography-tandem mass spectrometry (LC-MS/MS) methods. The first method included cortisone (E), F, 11-deoxycortisol (11 S), 17αOH-progesterone (17αOH-P4), progesterone (P4), dehydroepiandrosterone (DHEA), androstenedione (A), and T [23], while the second included estrone (E1), E2, DHT, and 17αOH-pregnenolone (17αOH-P5) [24]. Four serum samples were analyzed for each of the 21 eligible subjects, and all samples were measured in two consecutive runs. Due to analytical specificity requirements, as previously detailed [23, 24], data were discarded relating to 17αOH-P5 in five samples from four subjects, to P4 in three samples from one subject of E2 in three samples from three subjects, and to DHT in one sample. In addition, a very small serum volume was available for one sample, resulting in the sensitive measurement of only the T value.
We measured plasma CBG using the Biovendor ELISA kit (RD192234200R; Brno, Czech Republic). Free cortisol (free-F) was measured by the law of mass action according to the formula free-F (nmol/l) = 1000 * ((Z2 + 0.0122 C)0.5–Z), where Z = 0.0167 + 0.182 × (F-CBG), with F and CBG in μmol/l [25, 26].
Steroid enzymatic pathways
We analyzed the following steroid product-over-precursor molar ratios as indirect surrogate measures of the respective enzymatic activities: a) A/17αOH-P4 and DHEA/17αOH-P5 ratios for 17α-hydroxylase/17,20-lyase (CYP17A1) in Δ4 or Δ5 pathway, respectively; b) 17αOH-P4/P4 ratio for 17α-hydroxylase (CYP17A1); c) A/ DHEA and 17αOH-P4/17αOH-P5 ratios for 3β-hydroxysteroid dehydrogenase type 2 (HSD3B2); d) T/A and E2/E1 ratios for 17β-hydroxysteroid dehydrogenase type 3 (HSD17B3) and type 1 (HSD17B1), respectively; e) DHT/T ratio for 5α-reductase (SRD5A); f) E1/A and E2/T ratios for aromatase (CYP19A1); g) 11S/17αOH-P4 ratio for 21-hydroxylase (CYP21A2); h) F/11 S ratio for 11β-hydroxylase (CYP11B1), and i) F/E for 11β-hydroxysteroid dehydrogenase type 2 (11βHDS2) [20].
Steroid levels
Pre-treatment steroid profile
There were no significant differences in basal estrogen levels (E1 and E2), glucocorticoids (F, E, and 11 S), androgens (T, DHT, DHEA, and A), 17αOH-P5, CBG, and free-F in the basal samples collected before CC and Plac treatments (▶Tables 1 and ▶2). This highlighted the good efficacy of the drug wash-out period for these hormones. However, the serum 17αOH-P4 level before CC was significantly higher than before the Plac treatment (▶Table 1).
Effects of the CC and Plac treatments on steroid levels
▶Table 3 shows the fold changes in steroid levels induced by CC and Plac at the end of the 3-month treatment. The CC therapy increased serum P4 (p=0.044), 17αOH-P4 (p<0.001), A (p=0.025), T (p < 0.001), DHT (p < 0.001), E1 (p < 0.001), E2 (p < 0.001), F (p<0.001), and E (p<0.001), while the Plac treatment did not significantly change the serum corticosteroids, estrogens and androgens, except for an increase in 17αOH-P5 (p=0.039) and T (p=0.030) levels. In addition, CBG (p< 0.001), but not free-F, increased significantly during the CC treatment (▶Table 2). When the fold changes induced by CC were compared with those observed after placebo, we found that the effect of CC treatment on T, E2 (both p < 0.001), 17αOH-P4, DHT (both p = 0.001), E1, F (both p = 0.002) and E (p = 0.028) was significantly higher.
Steroid product-over-precursor ratios
Profile of pre-treatment product-over-precursor ratios
Steroid product-over-precursor ratios observed before the CC treatment were not significantly different from those observed before the Plac treatment, except for the CYP17A1 activity (A/17αOHP4), which was significantly lower before CC than before the Plac (p = 0.023).
Effects of the CC and Plac treatments on steroid productover-precursor ratios
▶Table 4 shows the effects of both CC and Placebo treatments on the steroid product-over-precursor ratios. CC significantly increased CYP17A1 (17αOH-P4/P4, A/17αOH-P4), HSD17B3(T/A), HSD17B1(E2/E1) (all p < 0.001), HSD3B2 (17αOH-P4/17αOH-P5), 11βHDS2 (F/E) ,(both p=0.001), HSD3B2 (17αOH-P4/17αOH-P5, p=0.001; A/DHEA p=0.005), CYP19A1 (E1/A, p=0.006), CYP21A2 (11S/17αOH-P4, p=0.002) and CYP11B1 (F/11S, p<0.001) activities, whereas the Plac did not significantly affect any of the analyzed ratios. In particular, the effect of CC treatment on CYP17A1 (17αOHP4/P4, p<0.001; E2/E1, p=0.001), HSD3B2 (17αOH-P4/17αOH-P5, p = 0.009), HSD17B3 (A/17αOH-P4, p = 0.004), HSD17B1 (T/ AP=0.003), CYP19A1 (E1/A, p=0.021), CYP11B1 (F/11S, p=0.008), and 11βHDS2 (F/E, p=0.023) was significantly higher than the effect of Plac.
Discussion
It is generally accepted that CC, a nonsteroidal estrogen agonist/ antagonist belonging to the SERM family, inhibits hypothalamic estrogen feedback, thereby increasing gonadotropin-releasing hormone (GnRH) production and stimulating gonadotropin secretion, which, in turn, results in increasing serum T levels by stimulating the Leydig cells [27]. To date, a larger number of studies have confirmed this CC effect on testicular steroidogenesis [28]. In fact, CC improves serum levels of LH and T in patients affected by a functional deficit of T, as it does in dysmetabolic conditions (i. e., T2DM and/or obesity) [5,6] with and without infertility [29], also in young or elderly men [30] and those affected by erectile dysfunction [31], in the absence of severe organic testicular damage [2].
CC could therefore be an alternative and probably more appropriate treatment than testosterone replacement therapy in patients affected by functional hypogonadism and high cardiovascular risk such as dysmetabolic men [32, 33].
Notably, evidence on the effects of CC on the overall testicular steroidogenesis is well established, yet the evidence on its effect on adrenal steroidogenesis is scant even though it has been demonstrated that both glands express the LH and estrogen receptors and could be affected by CC treatment [34–36]. To examine the enzymatic pathways behind the CC therapeutic effect, we thus extended our steroid analysis to a broad serum profile, as facilitated by the state-of-the-art LC-MS/MS technique, in the study cohort previously investigated in the CC therapy-controlled study [5]. We also measured CBG concentrations to assess the potential CC effect on both serum total F and its free fraction, which is known as the active hormonal form and the reliable expression of adrenocortical function [37].
We first ascertained that patients exhibited similar baseline circulating steroid levels before undergoing each treatment. Only 17αOH-P4 was higher before the CC treatment in comparison with the Plac phase. However, we found no explanation for such an isolated hormone increase. To explore the possible reason for this observation, we analyzed the carry-over effect between the two interventions in the two arms and found no significant impact (data not shown). Our data thus led us to attribute the difference in 17αOH-P4 basal levels to chance.
*In our previous studies, we showed that, besides T, CC therapy results in increasing serum DHT, E2, the sex hormones binding globulin (SHBG), LH, and FSH [5,38]. In the present study, we added the novel finding that CC administration along with Met significantly increases circulating E1, 17αOH-P4, F, E, and CBG, but not other steroids including androgen and glucocorticoid precursors, such as 17αOH-P5, A, DHEA, and 11 S levels (see ▶Table 3).
Notably, the sole administration of Met had no impact on the hormone levels examined. The CC-induced CBG increase was expected and potentially explained both by the estrogen stimulation of hepatic protein production [15] and by CC direct estrogen-like activity [39]. Moreover, the increase in CBG levels (approximately 30%) did not exceed the normal range (300–700 nmol/l) [40] and the calculated free-F levels were unchanged. These data, in addition to the unchanged adrenal androgen (DHEA, A, and 17αOHP5) and corticosteroid (11S) blood levels observed during CC treatment, suggested a lack of any direct adrenal stimulation by this drug [8]
To understand which pathway was affected the most by the CC treatment, we analyzed the molar ratios derived from the product-over-precursor steroid pairs as an indirect estimate of the corresponding enzyme activity in androgen and glucocorticoid synthesis, as well as in their metabolism in target tissues. In our cohort, CC therapy combined with Met significantly increased A/DHEA and T/A ratios, while reducing the A/17αOH-P4 ratio but not the DHEA/17αOH-P5 (see ▶Table 4). These data suggest a possible increase in the enzyme activity of 3β-hydroxysteroid dehydrogenase type 2 and 17β-hydroxysteroid dehydrogenase type 3 and an unclear suppression of the 17α-hydroxylase/17,20-lyase activity. The possible steroid pathway CC-induced is summarized at ▶Fig. 1.
As the adrenal glands are rich in the 17,20-lyase but poor in the 17β-hydroxysteroid dehydrogenase type 3 enzyme [20, 41], our data point to the CC effect at the testicular level, thereby indicating that the main effect is on the Δ5 pathway, rather than the Δ4 pathway for androgen synthesis [42]. Interestingly, our data indicate that CC therapy seems to act on testis steroidogenesis in a similar way to the hypothesis put forward for human chorionic gonadotropin (hCG). In fact, hCG stimulation in healthy and fertile males (doses ranging from 1500 UI to 5000 UI im) appears to enhance T secretion above all by facilitating the Δ5 pathway, due to the reduction in CYP17 activity (17α-hydroxylase/17,20-lyase) (Δ4) [43, 44].
In our patients, the CC therapy succeeded in increasing estrogens (E1 and E2), DHT, and the CYP19A1 (E1/A) activities, whereas the SRD5A (DHT/T) remained unchanged, thus suggesting that aromatase and 5α-reductase are only partially stimulated or unaffected by CC, respectively.
Our results showing the stimulating effects of CC therapy on aromatase activity agree with previous data obtained in oophorectomized women [9] and normal men [8]. In addition, we showed that CC elicited a higher than two-fold increase in serum E2. This could be due to the reduced hormone excretion, given a concomitant increase in serum SHBG by CC in our patients [5]. However, the affinity of E2 for SHBG is much lower than in the active androgens T and DHT (30% and 50%, respectively) [45], which are also increased by CC treatment, thus suggesting that the reduced excretion plays a limited role in the serum total E2 increase. On the other hand, the peripheral conversion of T into E2 is known to explain up to 40% of E2 levels [46], which would suggest only a limited impact of peripheral aromatase on the E2 level observed upon CC treatment. Therefore, the main source of the CC-induced increase in total serum E2 in our patients may, in fact, be the aromatase activity at the Leydig cell level [47, 48]. A reduced ratio between circulating T and E2 levels has been associated with damage in both the Leydig and Sertoli cells (primary hypogonadism) as well as with the failure of aromatase inhibitors or anti-estrogens, such as CC, to improve semen parameters and T levels [49,50]. We found that CC treatment induced a parallel increase in circulating T and E2 without altering their balance. In fact, these results highlight that there was no impairment in testicular steroidogenesis in our patients.
The increase in DHT levels that we observed after CC therapy does not appear to involve an increased 5α-reductase activity, as the DHT/T ratio was unchanged. In addition, the conversion rate of T into DHT by 5α-reductase has been reported to be 3.5 % [51]. Given such a conversion rate and considering the serum T levels obtained after the CC therapy (18.7 nmol/l), the estimated amount of DHT that should derive from T conversion would be around 0.7 nmol/l, which is twice the increase in serum DHT that we found at the end of the CC treatment.
These data also enable us to exclude the activation of the peripheral backdoor pathway to DHT, although the rise in serum 17OHP4 (40 %) could facilitate its peripheral metabolism [20]. Therefore, the increase in DHT level that we observed after CC therapy is probably not the result of an increase in its synthesis but is rather due to the reduction in its metabolic clearance, as supported by the previously mentioned increase in SHBG levels [5]
Finally, CC treatment increased the serum levels of 17αOH-P4, E, and F by 1.2–1.6 fold. In particular, 3-month CC therapy increased serum F levels (425 ± 113 nmol/l) compared to the basal levels (275±105 nmol/l), but these F levels are still below the upper reference limit of the same assay (550 nmol/l) [22]. This was explained by the increase in CBG as shown by the unchanged calculated values of the free-F during CC treatment. A similar mechanism would also explain the increased values of E and 17OH-P4. These data are in line with a few studies on the CC effect on CBG which showed an increase in the protein related to the treatment through an estrogen-like effect [10,38]. However, we cannot completely rule out a synthesis by the adrenal gland, as apparently suggested by the CYP11B1 expression (F/11S), nor a reduction in the oxidative metabolism at the target tissues, as suggested by the reduced 11βHDS2 activity (F/E). In addition, the CYP21A2 (11S/17αOH-P4) was reduced, thus suggesting a possible compensatory protective mechanism against glucocorticoid excess induced by CC. On the other hand, it is very unlikely that there was a CC stimulation effect mediated by the increase in LH and E2 levels and the action on their respective receptors expressed at this site [36].
*To address these issues, determining the ACTH before and after each treatment phase would have been decisive, but unfortunately was not conducted and is thus the main limitation of our study.
In conclusion, in previous work, we showed that CC treatment improves the function of the HPG axis in obese dysmetabolic men with functional hypogonadism by increasing serum LH and, in turn, T and E2 [5]. In the present study, we found that the improvement in T secretion due to CC seems to be caused by the testicular Δ5 rather than the Δ4 pathway, without causing a significant increase in 5α-reductase activity at the target tissues. In addition, we showed that there is an enhancement of plasma F levels, but not of the free-F, indicating a lack of a direct effect of CC therapy on the adrenal gland.
Future studies are needed to establish whether these hormonal changes induced by CC remain or progress during longer treatment periods, which could counter the evident beneficial effects of CC treatment on testicular steroidogenesis.
