Clinical efficacy of recombinant gonadotrophins

The advent of recombinantly-derived gonadotrophin preparations has been heralded as a major breakthrough in the therapeutic armory of assisted reproduction treatments (McDonough, 1995). Godwin Meniru poses a timely question by asking whether the new generation of drugs is providing any real benefit (Meniru, 1999). To date there are two recombinant follicle stimulating hormone (rFSH) preparations: follitropin α (Gonal-F; Serono Laboratories) and follitropin β (Puregon, Organon Laboratories). In discussing the benefits of a gonadotrophin preparation, one has to consider clinical efficacy, side-effects and cost-effectiveness. Clinical efficacy includes the ability to stimulate folliculogenesis, the production of mature oocytes, appropriate steroidogenesis for endometrial development and, in the context of in-vitro fertilization (IVF), sufficient quality pre-embryos and, ultimately, good pregnancy rates. Before moving on to these fundamental issues, we would like first to discuss the particular points raised by the first paper in this debate (Meniru, 1999). Meniru (1999) cites another paper (Jacob et al., 1998) and uses it to damp down the enthusiastic initial industry-sponsored reports on the efficacy of rFSH preparations. We would suggest, however, that Jacob's paper should not have been published. Jacob et al. (1998) retrospectively analysed the last 191 women who had received human menopausal gonadotrophin (HMG) in their IVF programme and compared them with the first 162 women to receive rFSH. Different urinary gonadotrophin preparations were used and some patients received growth hormone in addition; different starting doses were also used depending upon baseline serum FSH concentrations and response in previous cycles. Furthermore, 150 IU rFSH was considered to be equivalent to 225 IU HMG, based on the recommendations of Out et al. (1995), even though the study by Out did not set out to demonstrate equivalence. Suffice it to say therefore that a retrospective analysis of a heterogeneous data set would be considered by many to be meaningless and does not really warrant further consideration. Meniru himself (1999) uses data from 13 patients given Puregon at different starting doses and reports differing responses and variable requirements to either increase or decrease the dose. He also states that six patients found the Puregon injections to be more painful than Profasi (human chorionic gonadotrophin; HCG), although this was not studied in any formalized way. Such anecdotal reporting is unscientific and, frankly, unhelpful. If we now move away from the examples cited by Meniru and discuss the broader issues there are indeed a number of points that are worthy of debate. The original sources of gonadotrophins for therapeutic use were post-mortem pituitary extracts and the urine of post-menopausal women. The former source was withdrawn because of cases of Creutzfeld–Jakob disease, which occurred predominantly in Australia, but also in Europe. The extraction and purification of post-menopausal urine was pioneered in Italy in the late 1940s to result in the production of HMG (Donini et al., 1949). A total of 20–30 l of post-menopausal urine was required to treat one patient with one cycle of HMG. Through the 1960s the extraction process to remove non-specific co-purified proteins became more sophisticated such that activity was increased 10-fold over the early preparations to 100–150 IU FSH/mg protein. Greater purity produced fewer hypersensitivity reactions and less discomfort from the smaller volume of the injection. Despite the vastly increased purity of HMG (menotropin) and uFSH (urofollitropin) compared with the original preparations, their active ingredients only made up 1–2 % of the final product. They still contain large amounts of urinary protein (including cytokines, growth factors, transferrins and other proteins that might modulate ovarian activity) which makes uniform standardization very difficult, leads to local reactions at the injection sites and occasionally systemic illness. The advent of monoclonal antibodies in the 1980s, enabled further purification to be achieved by specifically selecting FSH out from the bulk HMG (Howles et al., 1992). The extract was 95% pure with a several hundred fold enhancement of specific gonadotrophin bioactivity and known as `highly purified urinary FSH' (u-hFSH HP, Metrodin HP®; Serono Laboratories, Aubonne, Switzerland). Extended clinical trials comparing uFSH (urofollitropin) and highly purified FSH demonstrated equivalent ovulation and pregnancy rates. Much reduced hypersensitivity was reported such that the subcutaneous route could be adopted for administration (Dore et al., 1994; Howles et al., 1994; Wickland et al., 1994). However, the problems of supply, collection, transport, storage and processing of an ever-increasing requirement of urine remained and the pharmaceutical companies began to explore the emerging new technology of genetic engineering. The two genes for the α and β subunits of FSH were incorporated into vectors for introduction into cells from the Chinese hamster ovary cell line and hereby began the process that should, theoretically result in an unlimited supply (Hayden et al., 1999). It is the degree of glycosylation that affects the biological activity of FSH (Rose et al., 1999). This can only be measured by bioassay and is not measurable by immunoassay. Pharmacopoeial monographs taking account of the inherent precision of the methods in bioassays allow 95% confidence limits of 80–125% of the stated dose on estimates of activity, i.e. 60–94 units of activity in a 75 unit ampoule (a potential variation of up to 57% between ampoules from different batches). The same pharmacopeial requirements apply to the recombinantly-derived FSH preparations. There is evidence that there is heterogeneity between the different recombinantly-derived preparations, hence the nomenclature `follitropin α and β'. The relationship of isoform composition to function has been recently reviewed (Rose et al., 1999). Data from in-vivo bioassays suggests that one of the major factors which controls FSH action is the relative degree of clearance of different isoforms. It is interesting to note that those forms of FSH which are most potent in vitro, tend to be least potent in vivo (Rose et al., 1999). A large number of intrinsic and extrinsic factors affect the performance of a drug in vivo. In the case of rFSH, the pattern of glycosylation, specifically terminal sialylation of the protein backbone, has excited a great deal of interest, as it is crucial to the bioactivity of the hormone. Overall the isohormone composition of rFSH has proved to be very similar to pituitary extract but great effort has been spent establishing which forms have greatest bioactivity (Lambert et al., 1995) in order to design the most specific and predictable drug. Sialylation determines acidity and isoelectric charge. Basic forms have higher receptor binding activity and intrinsic bioactivity but are cleared more rapidly than acidic forms. The more acidic isoforms have a 20-fold higher in-vivo bioactivity, mainly due to their higher absorption, lower clearance rate and longer elimination half-life. The pharmacokinetics of rFSH (Gonal-F®) are very similar to uFSH (Metrodin®) (Le Cotonnec et al., 1994a,b). In the future it is not unreasonable to foresee modifications to the molecular structure that lead to an extension of the half-life and in-vivo bioactivity. This could enable the frequency of injections to be reduced, which would be greatly appreciated by the patient. There are a number of immediately apparent advantages of rFSH over its urinary predecessors. Aside from the improved logistics of the pharmaceutical process, controlled manufacture has undoubtedly led to a more homogeneous product with much reduced inter-batch variability compared with the purification of enormous quantities of heterogeneous urine (Loumaye et al., 1995). The supply is potentially unlimited and shortages should no longer be a threat to clinical practice. There is no risk of infection or contamination with drugs or their metabolites as there is with products from a human source. The manufacturers have also confirmed that there have been no reported cases of seroconversion to anti-gonadotrophin antibodies (Recombinant Human FSH Study Group, 1995; Out et al., 1996). The purity of the products has certainly enhanced their administration which is effective, safe and much less traumatic when the s.c. route is adopted (Out et al., 1995; Albano et al., 1996). The most obvious advantages of rFSH are greater purity and specificity. It has been inferred that smaller doses and a more predictable response will result (Bergh et al., 1997). The most dramatic disadvantage to the health provider and patient however, is a marked increase in the price of the product. And herein lies the biggest question: are the recombinant preparations more cost effective? Research to date has been focused on comparing follitropin α (Gonal-F®; Serono) follitropin β (Puregon®; Organon) with urinary preparations of FSH. They have not been compared with each other directly, although any differences that might exist between them are likely to be subtle. Clinical evidence followed in therapeutic trials for IVF suggesting that rFSH (Puregon®) yields more oocytes, embryos and on-going pregnancies with a smaller dose for a shorter time than uFSH (Metrodin®) (Out et al., 1995). Frozen–thawed embryos were included in the figures as they resulted from the stimulation cycle involving the gonadotrophin in question. On-going pregnancy rates (at 12 weeks gestation) were consistently higher in the rFSH treated patients but not significantly so. Embryo quality appeared to be improved and having a surplus of embryos from which to choose the best to transfer increased the number available to be frozen. It is interesting to note that an initial comparative study of rFSH (Gonal-F®) with uFSH failed to show a significant difference in the number of retrieved or fertilized oocytes (Recombinant Human FSH study Group, 1995). The duration of treatment and the average dose required were also similar as were the numbers of patients reaching embryo transfer and implantation (clinical pregnancy). However, a later and larger comparative study of rFSH (Gonal-F®) with uFSH HP (Metrodin HP®) did reveal significant differences in the number of oocytes recovered (increased) and the number of FSH treatment days (reduced) (Bergh et al., 1997). The study included patients undergoing intracytoplasmic sperm injection (ICSI) where oocyte maturation is assessed prior to injection. There was no significant difference in oocyte maturity in the two groups. The clinical pregnancy rate for those patients who reached embryo transfer was similar for both groups suggesting no difference in embryo quality. This is interesting because pregnancy rates tend to be greater if there is a larger embryo pool from which to select for transfer (Staessen et al., 1993). Because of the suggestion of increased potency of rFSH, NV Organon have chosen to market Puregon® in 50 and 100 IU ampoules (rather than the 75 IU and 150 IU previously chosen). Using multiple or fractions of ampoules is commonplace and contributes to the potential inaccuracy of the actual dose administered (especially in the inexperienced hands of the self-injecting patient). No trials have yet shown conclusively that 50 IU Puregon® is equivalent to 75 IU rFSH (Gonal F®) or uFSH. Furthermore it is essential to avoid confusion in prescribing practices and refer to doses in terms of IU and not in terms of multiples of ampoules. Patients with polycystic ovaries certainly benefit from a reduced starting dose because of their propensity to an exuberant response to stimulation and increased risk of ovarian hyperstimulation syndrome (OHSS) (MacDougall et al., 1993). In the studies of rFSH to date, no significant differences in the incidence OHSS were found; however, since rFSH appears to be more potent, patients may be at greater risk thus vigilant monitoring remains essential in preventing this potentially fatal condition. Fractions of ampoules have been employed for both ovulation stimulation regimens and also ovulation induction for anovulatory infertility. It can be difficult to accurately dilute the contents of an ampoule to provide a precise dose when using fractions (e.g. 1.5 75 IU ampoules should provide 112.5 IU) and the ability to start with doses of either 50 IU or 100 IU may be advantageous for patients with polycystic ovaries (PCO) (Devroey et al., 1998; Hayden et al., 1999a). Ultimately the pharmaceutical companies are planning to move away from lyophilized preparations to ampoules containing solutions that can be better titrated to the individual's needs by giving more precise incremental doses (H.J.Out, personal communication). Such formulations might certainly prove to be advantageous to the patient prone to overstimulation. The advantage here of the recombinant preparations is their stability in solution and this is a significant advance over the urinary preparations which, whilst maintaining immunoactivity over time, have been shown to have declining bioactivity in storage (Braileanu et al., 1998). Very few studies have been performed comparing rFSH with menotropin, HMG which contains urinary contaminants and high concentrations of luteinizing hormone (LH). Favourable comparisons in terms of numbers of oocytes retrieved and on-going pregnancy rates with rFSH were demonstrated (Jansen and Van Os, 1996) but the study numbers were too small to reach statistical significance (89 women treated). However, a meta-analysis of urinary FSH with HMG has demonstrated that a significantly higher clinical pregnancy rate appears to be achieved with uFSH (Daya et al., 1995). This study implied that an adequate concentration of endogenous LH exists to achieve follicular and endometrial maturation, despite down-regulation of the pituitary with a gonadotrophin-releasing hormone (GnRH) analogue. Moreover, it has been suggested that exogenous LH supplementation in the form of HMG may be detrimental to the chances of achieving a pregnancy (Loumaye et al., 1996). If, for the it appears that rFSH is better than should it be to the The here is cost-effectiveness. In the and some other the urinary preparations are out to for recombinant FSH. The of the preparations have increased There are a number of new in the pharmaceutical the most in the with their HMG The of the drugs has led to many using with data of its efficacy, although it is certainly clinical that the drug We therefore that there is a for a study of HMG rFSH. If one significant at with a based on a of that a difference of between a clinical pregnancy rate of is not significant we have that patients would be required in each of the study to prove the i.e. that no difference exists between two preparations. We are a study of this in order to this very clinical question. has been to clinical studies of patients undergoing ovarian hyperstimulation for are also used for the induction of ovulation of a very small number of in patients with but as for there have been a few that rFSH It has been demonstrated that of LH is not required to achieve follicular growth (Hayden et al., and LH does not the FSH required to follicular growth et al., 1993). The of LH in the of is a of LH only in but has been with increased and rates in this et al., 1993). trials are still as to whether the of LH the in these The LH of HMG does not significantly increase serum LH to concentrations 1999). The FSH preparations, or those with a reduced LH however, no therapeutic advantage over HMG as the LH in HMG is compared with the endogenous of LH et al., et al., et al., et al., 1992). Furthermore, studies not a significant difference when HMG and FSH are administered et al., It is also that serum LH concentrations in response to as the follicle although some women with to LH in the of follicular growth a that may be due to production of ovarian factors in these patients and 1994). The that cells FSH and LH in order to in the ovary has been some patients with will to FSH et al., the will not et al., Furthermore, even when a follicle does in response to FSH it a lower concentration of and a for these patients preparations are In order to the of and multiple pregnancy in patients with the regimens and and have been by either regimens et al., et al., et al., et al., or regimens et al., 1993). studies have that of follicular growth a in the dose of exogenous FSH and the with follicular due to an increased number of FSH that the concentration of FSH required to growth is less than that required to The process of ovarian stimulation needs to be more controlled than for IVF as it is not to too many if high order multiple pregnancy is to be It has been suggested that the greater purity and of rFSH may to and ovarian response more to a greater of in terms of ovulation and fewer due to It can be difficult to the response to stimulation of a women with polycystic ovaries indeed this is the greatest therapeutic in ovulation induction The polycystic ovary is at least when by an exuberant and response to To date, there has been evidence to that rFSH is significantly better than uFSH in terms of pregnancy rates but rFSH (Puregon®) appears to be more uFSH in terms of requirement and duration of stimulation in ovulation induction et al., 1998). We have recently the first study of the use of rFSH (Puregon®) in the induction of ovulation in patients with polycystic ovary syndrome (Hayden et al., and demonstrated that rFSH can be used to stimulate follicular growth at a starting dose of 50 of is still required and the advent of rFSH in solution might further enable of dose The of genetic will have much more to in the to LH should enable a more to be administered than the LH will also in ovulation generation for the patient with et al., LH Study Group, 1998). In a new follitropin with a half-life by recombinant technology was reported et al., 1992). The longer half-life of compared to LH was as a function of the and at the terminal which did not with the LH receptor but did lead to in the and therefore a longer biological The of the was and The terminal of was and to the of the The new was into a Chinese hamster ovary cell line and the protein have confirmed a longer This is to recombinant which can now be to the patient with of and preparations. precise gonadotrophin action will be to the such that will be to to specific such as in the follicle or 1995). It could be that there is a case to be made for the use of gonadotrophin preparations or them out If we in the for gonadotrophin we can an for The most is which pregnancy rates equivalent to gonadotrophin treatment et al., 1994). is used for cases and appears as as ovarian et al., such as might also have a to as might the in-vitro maturation of oocytes 1999). The use of gonadotrophins to stimulate the ovaries has certainly the of The of gonadotrophin treatment regimens includes ovarian hyperstimulation syndrome and the risk of ovarian et al., 1998). in the might cycle IVF with the of one or two oocytes, embryo to and transfer a real for the studies of embryos might enable to and select those that are most likely to and 1999). like have been the of the in-vitro maturation of oocytes from or ovaries et al., 1998). as in the IVF and with the gonadotrophin preparations, there have been significant in the clinical of assisted are to to the patient and 1998). such as using an for weeks prior to starting a and pituitary the for and the risk of et al., 1998). has the of monitoring both the ovarian and the endometrial response to the for of 1994; et al., 1994). The introduction of should further the of women treatment by first the side-effects of by the and the of the cycle by to gonadotrophin in the early follicular and preventing an LH with administration of either a or multiple doses of an et al., 1998; Study Group, 1998). The total dose of gonadotrophins used also appears to be reduced when compared with a et al., 1995). Furthermore, can also be used in to the rate of cycle and thus allow assisted treatment in the treatments are to gonadotrophin stimulation they should not be if we are to further to the and of assisted reproduction it to that the the of the studies that the use of gonadotrophins and the regimens have been by the pharmaceutical who have away from to is the of most forms of and to the cost of the treatment of assisted reproduction it is that the recombinant preparations will be produced in sufficient quantities to enable the to be reduced have to be in terms of the of treatment with a rate of and the of with no health to their now as when first in a debate in this we stated that whilst ovulation induction a in women with induction on the other one of the least of therapeutic as the ovaries are in the for 1995; et al., 1996). many lead to the of many with the of some for later we have in the and further study of in-vitro maturation of oocytes, fertilization and implantation might lead to the transfer of a or to be is whether it is more for the oocyte and to have one or two of ovulation for IVF or a few more in which fewer are