Background
The twentieth century witnessed the discovery of pituitary gonadotropins follicle-stimulating hormone (FSH) and luteinising hormone (LH) that were made available as medication after the extraction and purification from the urine of menopausal women [
1]. The combination of urinary FSH (u-FSH) and human chorionic gonadotropin (u-hCG), a placental hormone displaying LH activity, has been available for the last forty years under the name of “human Menopausal Gonadotropin” (hMG) About twenty-five years ago, genetic engineering developed recombinant FSH (r-FSH) and recombinant LH (r-LH) by inserting the corresponding human DNA into Chinese hamster cells and then extracting and purifying the final molecules from their supernatant [
2,
3]. Nowadays a highly purified hMG (HP-hMG, Meropur, Ferring Germany) has been introduced.
Urinary-derived and recombinant gonadotropins have been widely used to treat women with infertility due to chronic anovulation [
4], or to provide a therapeutic stimulation to spermatogenesis [
5]. In most cases, however, these are administered to obtain the so-called Controlled Ovarian Stimulation (COS), that is the multiple follicular developments aimed at getting the number of oocytes needed to perform in vitro fertilization (IVF). Several studies, as well as systematic reviews, have compared the effectiveness of urinary gonadotropins (u-FSH or HP-hMG) vs. r-FSH for COS, showing the superiority of r-FSH over u-FSH [
4,
6,
7] and a substantial equivalence of r-FSH and HP-hMG [
8‐
13]. Also, only in few countries the combination of r-FSH + r-LH in a 2:1 ratio (Pergoveris, Merck, Germany) for the treatment of patients with hypogonadotropic hypogonadism [
14], is also licensed for COS. Comparative studies vs. hMG are rare and substantial do information are still missing [
15,
16]. To date, the largest study comparing r-FSH + r-LH vs. HP-hMG in human IVF was conducted retrospectively on real-life data from clinical practice that were obtained in the IVF Unit of a University Hospital and in a private IVF Clinic [
17]. In this context the r-FSH + r-LH association was demonstrated: (1) as effective as HP-hMG when the retrieved oocytes were less than 4, slightly (but not significantly); (2) superior when the retrieved oocytes were 5–8 and(3) significantly more effective in terms of pregnancy rate per embryo transfer (PR/ET) when they were 9 or more [
17]. Moreover, in the same study, the advantage of using the r-FSH + r-LH therapy was even more pronounced when only mature oocytes were considered [
17]. In support of the mentioned results, a multivariate logistic regression model confirmed that both the use of recombinant gonadotropins and the number of retrieved oocytes were increasing significantly the probability of a pregnancy, with an odds ratio (OR) of 1.628 and 1.083, respectively [
17].
In another paper, a comparative analysis of legal restrictions on access to IVF was conducted in 13 EU countries. This study demonstrated as countries with the most generous public financing scheme tended to restrict access to IVF to a greater degree. Contrarily, no link was established between IVF utilization and the manner in which coverage was regulated or the level of public financing was set [
18]. As a result of that, regulations seem generally more restrictive compared to the eligibility criteria in order to limit, through the reduced size of the covered population the budgetary outlays [
18].
Nowadays the cost of pharmacological therapies represent a major issue [
19] at international level and it is often considered as important as the effectiveness. Based on these premises, in the present study we aimed at performing a pharmaco economic analysis to estimate the cost-effectiveness of COS with r-FSH + r-LH in comparison with HP-hMG, considering the effectiveness of r-FSH + r-LH or HP-hMG (number of pregnancy, Positive or negative hCG test, Clinical pregnancy, Miscarriage, Cycle with embryo freezing, and Dropout) in terms of the Quality Adjusted Life Years (QALYs).
Discussion
The association of recombinant FSH plus recombinant LH in 2:1 ratio (Pergoveris, Merck, Germany) may be used not only to induce ovulation in anovulatory women with hypogonadotropic hypogonadism [
14], but also to achieve multiple follicular developments in human IVF. To date, there are still scarce or poorly informative data that compared the COS using r-FSH + r-LH with that obtained using other medications (e.g.HP-hMG) in the IVF setting. Indeed in several studies [
8‐
13] the urinary-derived HP-hMG was compared to r-FSH alone, without taking into account that HP-hMG contains not only FSH, but also LH and hCG that act on the same receptor of LH and have a powerful LH activity.
A randomized controlled trial (RCT) comparing HP-hMG and r-FSH + r-LH in patients undergoing IVF included only 122 women and showed comparable outcomes in terms of pregnancy rate [
16]. Unfortunately, this study was clearly underpowered as it was designed to assess the difference in oocyte rather than in the pregnancy rates. Another RCT study on 579 patients, included patients submitted to intrauterine insemination (IUI) rather than to IVF [
15]. This showed comparable outcomes, but without enough observations to show a significant difference of effectiveness between medication. In fact, IUI allows to get a much lower pregnancy rate than the expected when using IVF in patients of the same age (17.3% in the cited study [
15] vs. approximately 30–40%).
To the best of our knowledge, the wider study comparing HP-hMG and r-FSH + r-LH in IVF patients is a retrospective analysis of real-life data that included 848 women classified as expected “poor” or “normal” responders to gonadotropins [
17]. Data were collected under real-life practice circumstances in IVF Unit in S. Anna Hospital (Torino, Italy), and the pregnancy rate with fresh embryo transfer was calculated by stratifying patients according to the number of retrieved oocytes, in order to exclude that a difference in the results could be due to oocyte availability [
17]. The study showed an improvement in pregnancy rate according to the increasing number of retrieved oocytes. However, when comparing patients within the same oocyte, the two medications obtained comparable results when up to 4 oocytes, and a slight/ not significant advantage with 5–8 oocytes. However, the PR/ET became significantly higher when 9 or more oocytes were available [
17]. The multivariable logistic regression analysis confirmed that both the use of r-FSH + r-LH and the total number of retrieved oocytes increased the probability of pregnancy, with an odds ratio (OR) of 1.628 and 1.083, respectively, showing that the medication used for COS was even more influent than the number of oocytes itself.
Despite the contribution yielded by the previous literature in the field, the cost of COS, is still a major issue in times of global economic restrictions. This is relevant both in health systems where the patients cover the costs with their own resources and in countries where the National Health Service (NHS) takes care, partially or completely, of the expenses. Due to the evidenced gap of knowledge and the relevance of the proposed subject, the present study represents the first attempt to evaluate the cost-effectiveness of COS when recombinant gonadotropins in 2:1 combination (Pergoveris) or HP-hMG are used in a large series of patients undergoing IVF. Precisely we aimed to measure, through a sophisticated economic analysis the overall cost-effectiveness of IVF cycles based on the previously published database [
17].
As a result, the present analysis demonstrates that r-FSH + r-LH therapy showed higher cost-effectiveness than HP-hMG in the considered two-years observation periods with a slightly lower overall cost per pregnancy despite a higher cost per medication. The cost-effectiveness acceptability curve (CEAC) showed that the observed difference between the two medications was likely to further increase if the time horizon was prolonged and the number of IVF cycles rose. The advantage given by recombinant gonadotropins vs. HP-hMG was not linked to a higher number of retrieved oocytes because calculations were performed after patients’ stratification in subgroups and having the same number of available oocytes. In detail, for the population of patients obtaining 5 or more oocytes, the r-FSH + r-LH therapy resulted to be highly cost-effective compared with HP-hMG, with an ICER period equal to €11,365 per QALY. Interestingly the patients that obtained the highest advantage from being treated with r-FSH + r-LH instead of HP-hMG were those with 10–15 retrieved oocytes, who had an ICER of € 6569 per QALY. According to the medical literature, patients getting 10–15 retrieved oocytes are among those with the best prognosis after fresh embryo transfer [
26] and may be considered “normal responders” to gonadotropin stimulation, representing approximately 40–50% of the overall population undergoing IVF.
Our analysis, showed two main limitations: (a) the lack of reports on the quality of life of patients undergoing IVF did not allow an exact quantification of the utility measures used in the model, some of which were obtained considering the opinion of IVF experts; (b) the database on which our analysis was based did not report about embryo freezing and dropouts from the program. Even in this case, their incidence was estimated according to the opinion of IVF experts. However, the one-way sensitivity analyses allow to take this uncertainty into account, and therefore these two limits should not have affected the validity of our results. A final limitation, c) regards the generalisability of our results on other countries. This study was specifically settled on the Italian general practice and the model was populated with Italian costs and tariffs different than what happens in other European countries. However, the regulation in Europe demonstrates a difference in term of access to IVF [
18]. Contrarily, no link was established between IVF utilization and the manner in which coverage was regulated or the level of public financing was set [
18]. As a result of that, we can assume that the general management of these patients could be homogeneous around Europe and the model structure represents a general approach for other countries. Further analysis, could consider the same model structure adapting the specific economic parameters and evaluate the cost-effectiveness results for each country perspective. The National Healthcare Assistance in Italy covers the complete cost for gonadotropin treatment in IVF, provided that the patient is younger than 45 and her basal FSH circulating level is below 30 U/l. Not all countries have such a system, but in general, most healthcare systems help patients to face a percentage (variable in different countries) of the economic cost of IVF medications and procedure. Although the results found herein are not perfectly applicable to other countries due to such differences in the healthcare assistance reimbursement, the relative proportion of the cost of ovarian stimulation with recombinant FSH and LH vs. hMG is rather constant everywhere, and therefore the general concepts expressed in this study may be of interest even outside Italy.
The strength point of this analysis is that is based on real-world data [
17]; the transition probabilities used to perform the model have been obtained from an Italian study which collected data from clinical charts of IVF Unit in S. Anna Hospital (Torino, Italy).
Conclusions
In conclusion, the present cost-utility analysis demonstrated that the r-FSH + r-LH combination, although more expensive than HP-hMG when medication costs are considered, may be effectively used to obtain COS in IVF patients without increasing the overall costs for the patients or the NHS. On the contrary, the r-FSH + r-LH association allows getting slightly reduced costs for pregnancy, improved cost-effectiveness and quality of life, especially when the so-called “normal-responders”, who represent the majority of IVF patients, are considered.