Effect of letrozole on follicular fluid steroids concentrations in cancer patients undergoing oocyte cryopreservation

Abstract

Purpose

To investigate the impact of letrozole administration on follicular steroid hormones during controlled ovarian hyperstimulation for fertility preservation.

Methods

One hundred and nineteen women with cancer undergoing oocytes retrieval for fertility preservation were recruited. All women underwent ovarian hyperstimulation according to a random start protocol. Those with hormone-sensitive tumors also received letrozole, an aromatase inhibitor aimed at keeping peripheral estrogen levels low. At the time of oocytes retrieval, a sample of follicular fluid was collected and frozen. All samples were assayed concomitantly after thawing, by liquid chromatography tandem mass spectrometry. The concentration of 15 steroid hormones was determined and results were compared between women who did and did not receive letrozole.

Results

Fifty-two women were treated with letrozole, while 67 were not. Statistically significant differences emerged for 12 of the 15 tested steroids. They were the following: cortisol, 11-deoxycortisol, 21-deoxycortisol, dehydroepiandrosterone sulfate (DHEAS), dehydroepiandrosterone (DHEA), estradiol, androstenedione, testosterone, dihydrotestosterone (DHT), 17-hydroxyprogesterone, progesterone and corticosterone. The most striking differences were observed for testosterone that showed a more than 200-time increase in women receiving letrozole. Estradiol was conversely reduced to a third.

Conclusions

The endocrine microenvironment surrounding oocytes is markedly perturbed by the concomitant assumption of letrozole. Robust clinical evaluation is pressingly needed to rule out any detrimental effect on the chance of live birth with the use of these oocytes.

Introduction

The progress made in primary systemic therapy has led to a great increase in the relapse-free survival of young breast cancer patients [1, 2]. This has allowed clinicians to focus on aspects related to patients’ quality of life, such as the desire of motherhood. Moreover, it has been demonstrated that pregnancy in breast cancer survivors does not increase the relapse risk. On the contrary, it seems that it can even improve the overall survival rate [3].

However, oncological treatments have a great impact on female fertility. Chemotherapy, especially alkylating agents, may damage ovarian reserve up to premature ovarian failure [4]. Endocrine therapy, administered for several years in patients with hormone-sensitive disease, obliges to delay pregnancy seeking until reaching an age at which female fertility could be impaired [5–7].

Fertility preservation has become mandatory in all these patients, who are thus mainly offered with oocytes cryopreservation before initiating oncological treatments [8]. However, high estradiol levels, reached during controlled ovarian stimulation, have been a matter of concern because of their possible effect on the growth of hormone-sensitive tumors. For that reason, more than a decade ago, a new protocol has been introduced combining gonadotropins, for ovarian hyperstimulation, with letrozole with the aim of maintaining low peripheral estrogen levels [9, 10].

Letrozole is a selective non-steroidal aromatase inhibitor which blocks the conversion of androstenedione and testosterone to estrone and estradiol, respectively [11, 12]. However, considering its action on estrogens/androgens levels, we can assume that letrozole may induce extensive changes in the endocrine follicular environment, thus potentially influencing oocytes quality and competence.

Evidence regarding the possible detrimental effects of the concomitant administration of letrozole during ovarian hyperstimulation for oocytes cryopreservation is modest and controversial [13, 14]. Randomized controlled trials have not been performed yet and evidence obtained in non-oncological contexts, where letrozole is discontinued some days before ovulation trigger, is not informative. Indeed, for fertility preservation in cancer patients, the administration is continued up to some days after oocytes retrieval (extended use), thus exposing the whole process of folliculogenesis (including the final trigger) to its possible detrimental effects. Of relevance here is that some authors even claim that letrozole would not be necessary. There is evidence suggesting that fertility preservation in breast cancer patients would be safe and efficient even without the concomitant use of letrozole [15]. The levels of estrogen developed during ovarian hyperstimulation may actually be clinically negligible, regardless of the receptors status of the lesions [15]. The use of a therapeutic and potentially detrimental adjuvant in the absence of a definite demonstration of its capacity to protect patients from cancer spread may be questionable. Obtaining information on the quality of the oocytes retrieved after the extended use of letrozole is therefore fundamental. The best way to assess the competence of cryopreserved oocytes would be the evaluation of live birth rate: as long as these oocytes are not used, it is not possible to estimate their real quality. This implies that it would take several years to reliably evaluate this aspect (until a consistent number of women thawed and use their oocytes). Definitive data on pregnancy rate is not available yet. Meanwhile, indirect markers of oocytes quality, such as the endocrine microenvironment surrounding oocytes, may be taken into consideration. In this study, we aimed to provide evidence on this issue by comparing the steroids cascade in the follicular fluid in women undergoing oocytes cryopreservation for cancer and who did and did not receive letrozole.

Material and Methods

The present study is a single-center, biological, non-pharmacological, no-profit study conducted at the infertility unit of Fondazione IRCCS Ca’ Granda Ospedale Maggiore Policlinico, Milan, aiming at evaluating the concentration of 15 steroid hormones in the follicular fluids of women with cancer undergoing oocytes cryopreservation. Recruitment period lasted from January 2017 to December 2019. Laboratory analyses were carried out on scrap biological material: follicular fluids, aspirated during the oocyte retrieval and generally eliminated after oocytes separation, were used. Thus, patients’ participation in the study didn’t involve any interference with the normal clinical process. All the patients undergoing oocytes cryopreservation for oncological indications were considered for enrollment. Patients recruited were asked to sign an informed consent, after receiving a comprehensive explanation of the purposes and characteristics of the study. The study was approved by the local Ethical Committee.

Patients undergoing oocytes cryopreservation before oncological treatments followed a “random start” protocol, reported in detail elsewhere [16]. This implies the possibility to start ovarian stimulation in any phase of menstrual cycle, without waiting for the beginning of the next follicular phase. Briefly, women started treatment on the day of referral, irrespective of their menstrual cycle date, with long-acting recombinant FSH 100 or 150 µg according to body weight (< or ≥ 60 kg) (Elonva®, Merck Sharp & Dohme, UK) followed by recombinant FSH daily (Gonal-F®, Merck-Serono, Italy), if needed. Age was not considered in choosing the initial dose. Women were monitored with serial transvaginal ultrasounds and if required serum estrogen assessment, adding daily GnRH antagonists (Orgalutran® 0.25 mg Merck Sharp & Dohme, UK or Fyremadel® 0.25 mg Ferring, Switzerland) when the leading follicle reached the diameter of 13–14 mm up to the time of ovulation trigger. Final oocyte maturation was triggered with GnRH agonists (Fertipeptyl® 0.2 mg, Ferring, Switzerland) when at least three dominant follicles (mean diameter > 15 mm) were present, of whom at least one had a mean diameter ≥ 18 mm. Oocyte retrieval was performed under transvaginal ultrasound guidance 36 h after triggering. When less than 3 follicles developed, cycles were canceled, since in these situations (poor response) the balance of risks and benefits of oocytes retrieval was considered unfavorable. Only oocytes were preserved. None of the included women could cryopreserve embryos because this procedure is banned by law in Italy.

Women who were diagnosed with hormone-sensitive cancers were also given letrozole 5 mg (Letrozolo®, TEVA, Italy) daily for the whole duration of ovarian hyperstimulation and continued for 5–7 days after oocytes retrieval, in order to keep estrogen levels low in peripheral tissues. Breast cancers that did not express hormonal receptors were not given letrozole. Conversely, letrozole was given also to non-breast cancers thought to be potentially sensitive to sex steroids, such as in particular ovarian cancers. Patients’ participation in the study didn’t influence the choice of the stimulation protocol, which was decided based on the clinical condition. None of the included women received chemotherapy prior to hyperstimulation (this was an exclusion criterion for being included in our egg freezing program).

Laboratory analyses were carried out on scrap follicular fluid, aspirated during oocytes retrieval. For every enrolled patient, the pool of follicular fluids was collected, and then centrifugated at 2000 RPM for 10 min in 50-ml Falcon tubes at room temperature. The supernatant was collected, aliquoted in 1.5 ml Eppendorf and cryopreserved at -20 °C. The levels of 15 steroids [11-deoxycorticosterone, 11-deoxycortisol, 17-OH-progesterone, 21-deoxycortisol, aldosterone, androstenedione, corticosterone, cortisol, cortisone, dehydroepiandrosterone (DHEA), dehydroepiandrosterone sulfate (DHEAS), dihydrotestosterone (DHT), estradiol, progesterone, testosterone] were then assayed concomitantly after thawing.

Hormones were assessed by liquid chromatography tandem mass spectrometry (LC–MS/MS). The procedure is explained in detail elsewhere [17, 18]. Briefly, an IVD-MS steroids kit (MassChrom, Steroids in Serum/Plasma, Chromsystems, Gräfelfing, Germany) was used. Samples were prepared according to the manufacturer’s instructions: 500 µl of each follicular fluid sample, calibrators or quality control were placed in solid-phase extraction sample plate, previously equilibrated, with 50 µl of a deuterated internal standard mix solution and 450 µl extraction buffer. This mixed sample was then vortexed and centrifuged for 1 min at 400xg. The supernatant of the sample was evaporated under nitrogen to dryness, reconstituted with 100 µl of reconstitution buffer and two 40 µl aliquots were injected into the high-performance liquid chromatography system equipped with an analytical column (operating at 32 °C) for peak separation. Mobile phases, provided with the kit, were used for steroids elution. The chromatographic system was interfaced with a hybrid triple quadrupole/linear ion trap mass spectrometer (QTRAP 5500; Sciex, Monza, Italy) equipped with an electrospray ionization source, operating in positive mode. A blank calibrator matrix and six multilevel serum calibrators provided with the kit were used for calibration and three certified quality controls of serum were used to assess within- and between-run precision and accuracy. Analysis of duplicates was performed on 10% of the samples. Relative standard deviation was up to 10%. Accuracy, evaluated for low, intermediate and high-quality control, was in the range of 90–110%.

Data collected were transferred in Statistical Package for Social Science (SPSS 26.0, IL, USA) database for subsequent analyses. Differences between the two study groups were tested using Student t test or Mann–Whitney test or Fisher Exact test, as appropriate. Given the non-normal distribution of steroids hormones, these variables were compared using non-parametric statistics (Mann–Whitney test), and reported as median [interquartile range—IQR]. Sample size was calculated stating as clinically relevant demonstrating a threefold increase in the frequency of cases with steroids above the 90^th centile (or below the 10^th centile) in women receiving letrozole. Setting type I and II errors at the conventional 0.05 and 0.20, this corresponded to about 120 women (60 per group).

Results

Overall, 119 women were initially selected; 52 were treated with letrozole, while 67 did not. Baseline characteristics of the two groups are shown in Table 1. Women treated with letrozole showed older age and mildly higher BMI. Conversely, despite this difference in age, no differences emerged in ovarian reserve variables (antral follicle count—AFC, and Anti-Müllerian Hormone—AMH). As expected, indications to oocytes cryopreservation were different between the two groups, with letrozole being prescribed primarily to women with breast cancer (Table 1).

Table 1.

Baseline clinical characteristics of the study groups

Characteristics	With letrozole	Without letrozole	p
	n = 52	n = 67
Age (years)	36 [30—37]	30 [24—35]	< 0.001
BMI (Kg/m2)	22.5 [20.0—25.7]	21.3 [19.4—23.0]	0.05
Previous deliveries	4 (8%)	8 (12%)	1.00
Seeking pregnancy at the time of the diagnosis	6 (12%)	3 (5%)	0.18
Serum AMH (ng/ml)	2.41 [1.15—4.03]	2.22 [1.38—3.36]	0.99
Total AFC	16 [10-28]	18 [12-25]	0.47
Indication to oocytes cryopreservation			< 0.001
Hematological cancer	0 (0%)	39 (58%)
Breast cancer	48 (92%)	14 (21%)
Borderline ovarian tumors	2 (4%)	1 (2%)
Other ovarian tumors	2 (4%)	4 (6%)
Other cancers	0 (0%)	9 (13%) a

AFC: Antral Follicle Count. AMH: Anti-Mullerian hormone

^a Include melanoma (n = 2), colorectal cancer (n = 1), rhinopharyngeal cancer (n = 1), intracranial choriocarcinoma (n = 1), medulloblastoma (n = 1), thymoma (n = 1), multicystic mesothelioma (n = 1), pseudomyxoma peritonei (n = 1)

Data are reported as median [interquartile range] or number (percentage)

Ovarian hyperstimulation variables are shown in Table 2. The two study groups were homogeneous as regards the previous use of estrogen–progestins and the cycle phase at initiation of hyperstimulation. No differences emerged in the total dose of gonadotropins, duration of stimulation, number of developed follicles and number of oocytes retrieved. Unsurprisingly, serum estrogen levels at the time of trigger were significantly lower in the group taking letrozole, with concentrations three-to-four time lower. Conversely, serum progesterone levels were similar in the two groups (Table 2).

Table 2.

Baseline clinical characteristics of the study groups

Characteristics	With letrozole	Without letrozole	p
	n = 52	n = 67
Estroprogestins use at the time of recruitment	4 (8%)	9 (13%)	0.39
Cycle phase at initiation of hyper-stimulation			0.71
Follicular phase	32 (62%)	38 (57%)
Luteal phase	20 (38%)	29 (43%)
Total dose of gonadotropins (IU)	900 [675—1,350]	900 [600—1,400]	0.95
Duration of stimulation (days)	11 [9-11]	11 [10-12]	0.33
N. of developed follicles (diameter ≥ 11 mm)	17 [10-26]	18 [12-23]	0.39
Serum estrogens (pg/ml)	471 [266—731]	1,588 [932—2,990]	< 0.001
Serum progesterone (pg/ml)	405 [187—917]	569 [103—931]	0.43
N. of oocytes retrieved	13 [8-22]	14 [9-20]	0.94
N. of mature oocytes retrieved (frozen)	10 [6-14]	10 [6-18]	0.66
N. of mature oocytes per developed follicle	0.60 [0.49–0.80]	0.58 [0.36–0.79]	0.50

Data are reported as median [interquartile range] or number (percentage)

Follicular fluid levels of the 15 steroids analyzed are enlisted in Table 3. Among these, 12 steroid concentrations significantly differed between the two groups: cortisol, 11-deoxycortisol, 21-deoxycortisol, DHEAS, DHEA, estradiol, androstenedione, testosterone, DHT, 17-hydroxyprogesterone, progesterone and corticosterone. The most striking differences were observed for testosterone, androstenedione and DHT. Estradiol levels were reduced to a third in the group treated with letrozole (Table 3). For androstenedione, the IQRs did not even partially overlap. This is in line with the expected effects of letrozole (Fig. 1). These analyses were repeated including only women with breast cancer (n = 64) and comparing those who did (n = 48) and did not (n = 14) receive letrozole. Statistically significant differences were confirmed for 10 steroids, i.e., cortisol, 11-deoxycortisol, 21-deoxycortisol, DHEA, estradiol, androstenedione, testosterone, DHT, 17-hydroxyprogesterone and progesterone. Conversely, differences in corticosterone and DHEAS were no longer significant (data not shown).

Table 3.

Follicular steroids hormones in women who did and did not received letrozole

Steroid hormones	With letrozole	Without letrozole	p
	n = 52	n = 67
Progesterone (μg/l)	560.4 [465.9—747.7]	432.4 [334.1—615.1]	0.002
11-deoxycorticosterone (μg/l)	25.4 [20.9—36.1]	27.6 [21.6—35.2]	0.72
Corticosterone (μg/l)	3.1 [2.5—4.2]	2.5 [1.8—3.6]	0.028
Aldosterone (µg/l)	0.05 [0.02—0.10]	0.05 [0.02—0.09]	0.90
17-hydroxyprogesterone (μg/l)	618.0 [475.8—751.9]	432.5 [331.4—579.8]	< 0.001
21-deoxycortisol (μg/l)	0.21 [0.12—0.46]	0.14 [0.07—0.21]	0.002
11-deoxycortisol (μg/l)	1.27 [1.02—2.54]	0.78 [0.62—1.21]	< 0.001
Cortisol (μg/l)	49.5 [42.0—58.3]	58.0 [44.3—74.7]	0.04
Cortisone (μg/l)	13.3 [10.4—16.7]	13.3 [10.6—17.5]	0.75
DHEA (μg/l)	12.3 [7.8—17.1]	8.0 [5.6—13.4]	0.002
DHEAS (μg/l)	1,697 [1,061—2,104]	1,255 [969—1,659]	0.002
Androstenedione (μg/l)	153.2 [113.3—217.7]	4.3 [1.8—9.6]	< 0.001
Testosterone (μg/l)	23.9 [12.6—34.7]	0.10 [ 0.06—0.26]	< 0.001
DHT (μg/l)	0.27 [0.16—0.45]	0.11 [0.03—0.19]	< 0.001
Estradiol (μg/l)	104.3 [53.3—219.4]	246.1 [152.3—391.2]	< 0.001

DHEA: dehydroepiandrosterone. DHEAS: dehydroepiandrosterone sulfate

DHT: dihydrotestosterone

Steroids were grouped based on the main branches of the cascade (see Fig. 1)

Fig. 1.

Steroid hormone cascade. The hormones tested in this study are highlighted in yellow. Aromatase function within this cascade is highlighted in blue. DHEA: dehydroepiandrosterone. DHEAS: dehydroepiandrosterone sulfate. DHT: dihydrotestosterone

Finally, we correlated the rate of mature oocytes per developed follicle (a surrogate marker of the quality of the whole folliculogenesis process) with the 15 tested steroids, separately for the two studied groups. No significant correlation emerged in women who received letrozole. Conversely, in the non-treated group, a statistically significant positive correlation was found with the concentration of estradiol in μg/l (B = 0.01, 95%CI: 0.00–0.02, p = 0.01).

Discussion

At present, little is known about the quality of cryopreserved oocytes from women treated with letrozole for fertility preservation [13, 14, 19, 20]. Data on pregnancy rates with the use of these oocytes are yet scant. Oktay et al. reported on 131 women with breast cancer who underwent embryo cryopreservation program using letrozole: 33 (25%) returned to use embryos with a median time from cryopreservation to usage of 5 years. Seventeen (51%) had at least one child. This result was in line with the US national data [21]. Pereira et al. also showed that breast cancer patients undergoing ovarian stimulation with letrozole have live birth rates comparable to age-matched counterpart: 57 out of 129 women with breast cancer who underwent embryo cryopreservation proceeded with frozen embryo transfers, obtaining a live birth rate per cycle of 32%, in line with local results [13]. Of utmost relevance here is that, in both studies, embryos rather than oocytes were cryopreserved. Inferring evidence, obtained in embryos, to oocytes could be debatable. Oocytes are generally more vulnerable to freezing procedures than embryos and one may hypothesize that oocytes obtained with concomitant letrozole assumption could be less able to sustain the additional stress of freezing. In this regard, one should also emphasize that oocytes rather than embryo freezing is currently preferred, in order to preserve at best the reproductive freedom of choice of the woman [22].

Given the long follow-up periods needed to obtain robust data on the chances of live birth, a definitive evidence on the quality of oocytes with the use of letrozole will take years from now to be available. For this reason, we designed the present study using a surrogate mean to evaluate oocytes quality, i.e., by investigating the intrafollicular endocrine milieu. Twelve out of the 15 tested steroid hormones showed significantly different concentrations between women using letrozole and women undergoing a conventional ovarian hyperstimulation protocol. Particularly, testosterone, DHT, androstenedione and estrogen revealed the most striking differences. Of interest is that this de-arrangement was not limited to estrogens and androgens but involved also other branches of the steroidogenic cascade and in particular the DHEA-DHEAS arm and the 17-hydroxyprogesterone 21-alpha-metabolites up to cortisol. Only the metabolites derived from the 21-alpha-hydroxilation of progesterone were not affected (while progesterone itself was).

Before our contribution, three smaller studies analyzed follicular fluid composition in patients treated with letrozole. Garcia-Velasco et al. in 2005 evaluated serum and follicular fluid concentrations of four hormones (androstenedione, testosterone, estradiol and progesterone) and concluded that adding 2.5 mg of letrozole is associated with an increase of intraovarian androstenedione and testosterone concentration [23]. Two years later, Andersen and Lossl analyzed the impact of androgen priming on the follicular fluid concentrations of AMH, inibin-B, progesterone, androstenedione, testosterone and estradiol, comparing 15 women treated with 3 mg of cetrorelix alone (Group I), 15 women treated with cetrorelix plus 2.5 mg of Letrozole (Group II) and 15 women treated as Group II plus 1,200 UI of hCG (Group III). They showed how the increase in intrafollicular androgen levels due to letrozole use led to augmented granulosa cell production of AMH and reduced inibin-B production [24]. These two studies, however, included women who discontinued the use of letrozole before the ovulation triggering, a situation that could be different to what occurs in cancer women when the assumption has to be pursued up to the oocytes’ retrieval. Noteworthy, both studies highlighted only a less than twofold increase in testosterone and androstenedione and failed to show differences in estradiol [23, 24]. In this regard, the most informative contribution is the recent study from Goldrat et al. who compared follicular fluid levels of three steroids (estradiol, testosterone, progesterone) in 23 breast cancer patients treated with letrozole until the time of ovulation triggering and 24 infertile patients undergoing conventional hyperstimulation protocol [25]. They failed to detect differences in progesterone concentration, they showed decreased levels of estradiol, exclusively when ovulation was triggered with hCG (not with GnRH agonists), and they highlighted a tenfold increase in testosterone [25].

It is difficult to draw definitive conclusions on the possible effects of this drastic disruption of follicular fluid composition observed in our study on oocytes quality. However, in the non-letrozole group, we observed a positive correlation between the rate of mature oocytes retrieved per developed follicle and estradiol, suggesting but not demonstrating a potential positive effect of local estrogens. Goldrat et al. hypothesized that the increased testosterone/estradiol ratio could cause a suboptimal follicular fluid environment [25]. Previously, Lossl et al. demonstrated that high follicular fluid testosterone levels, due to the use of aromatase inhibitor, were associated with reduced oocytes fertilization rates but good embryo quality [26, 27]. They explained their finding assuming a dual effect of androgen on granulosa cell function: a positive effect on small antral follicles, consequent to the increase in FSH receptors and to steroidogenesis promotion, and a detrimental effect on later stages of follicular development. Of relevance here is that the magnitude of the differences was much more relevant in our samples obtained after a persistent use of letrozole. For instance, the median [IQR] follicular testosterone concentration was 23.9 [12.6—34.7] and 0.10 [ 0.06—0.26] μg/l in women who did and did not receive letrozole, respectively (corresponding to a more than 200-fold increase).

Drawing robust clinical inferences from our findings cannot be done. However, we believe that our data argues against the systematic use of letrozole for breast or ovarian cancer, even in cases of histological demonstration of the absence of steroids receptors on pathological specimens. When safety is concerned, cautious is warranted and this could justify the use of letrozole also in doubtful cases. On the other hand, women with cancer deeply rely on oocytes cryopreservation to the point of embarking in a burdensome additional procedure that, in some cases, can delay the initiation of oncological treatments. As physicians, we have the ethical duty to ensure the most effective options for these women, including, in our opinion, the wise and rational use of letrozole.

Compared to previous published studies, our study has greater sample size, allowing us to draw stronger conclusions. Moreover, the laboratory method used to assay follicular fluid steroid concentrations gives our data greater reliability: liquid chromatography tandem mass spectrometry (LC–MS/MS) is more precise and sensible, compared to other analytical methods available [17]. Besides, we did not limit our analyses to estradiol and progesterone concentrations, as other studies did before, but we considered all the molecules that take part in the steroidogenic cascade.

Some limitations of our study should be acknowledged. Firstly, the populations we compared were not homogeneous as regards baseline characteristics, thus exposing our study to some biases. The exclusive inclusion of women with breast cancer would have tempered this limitation, but would have also limited the statistical power and the precision of our study. Moreover, we did not have main reasons to foresee an important effect of the type of cancer on the composition of the follicular fluid. In addition, the large detected differences tend to rule out any relevant effect of this inaccuracy. Finally, it is noteworthy that even when restricting the analyses to the 62 women with breast cancer (48 with letrozole and 14 without), most statistically significant differences remained. Secondly, one must be generally cautious when interpreting follicular fluid steroid concentrations, since it has been demonstrated that they show high variability within and between subjects [28]. This occurred also in our study, as demonstrated by the very ample IQRs.

In conclusion, in no doubt the use of letrozole has a strong impact on follicular fluid steroids’ concentrations. Nevertheless, it is difficult to speculate on the effect of these changes on oocytes quality, especially considering that the most important data, i.e., pregnancy rate, are currently still missing. It is necessary to wait for these cryopreserved oocytes to be used, to better clarify this still unclear topic. In the meantime, we advocate for a wiser use of this drug in everyday clinical practice.

Funding sources

None.

Declarations

Financial disclosure statement

Edgardo SOMIGLIANA received during the last three years grants of research from Ferring and Merck-Serono. He also received honoraria from Theramex, Gedeon-Richter and Merck-Serono. All the other authors do not have any financial disclosure to declare.