A Multicenter Phase IV Study to Investigate the Immunogenicity of Recombinant Human Follicle-Stimulating Hormone and Its Impact on Clinical Outcomes in Females Undergoing Controlled Ovarian Stimulation.

Exogenous proteins, including therapeutic ones, are immunogenic and have the potential to cause antibody formation; the tendency of a protein to mount such a response is referred to as immunogenicity, which can be long-term and can lead to immunological memory. Development of such antibodies may lead to failure of a drug's development during clinical studies or may interfere with efficacy or lead to serious complications in the postmarketing phase.[ 1 2 ] Preparations of recombinant human follicle-stimulating hormone (r-hFSH) are used for controlled ovarian stimulation (COS) in infertile females prior to the performance of intrauterine insemination (IUI) or in vitro fertilization (IVF). FSH preparations are, in general, considered to have low immunogenic potential.[ 3 ] However, biological drugs are complex and variable in structure, and their manufacture involves complex biotechnological processes, making them quite sensitive to changes in manufacturing processes. Another contributing factor is that different manufacturers use different molecular clones and cell banks and may have different fermentation and purification processes.[ 4 ] Thus, different preparations of the same biological drug may vary in terms of purity, potency, and immunogenicity. The present study was envisaged as a prospective, multi-center clinical study to assess the immunogenicity of a r-hFSH preparation in patients with infertility, when used for COS as part of one, two, or three successive cycles of either IUI or IVF.

The study was conducted between September 26, 2014, and April 28, 2016. A total of 293 women were screened across the 12 sites, of whom 285 were enrolled and exposed to the study drug. Out of these, 180 were assigned to IUI, and 105 were assigned to IVF treatment. A total of 234 patients underwent one treatment cycle, while 33 and 18 patients underwent two and three treatment cycles, respectively. Thus, the total number of treatment cycles included in the study was 354. The mean (± SD) age and body mass index of patients were 30.1 ± 4.54 years and 23.1 ± 4 kg/m 2 , respectively. Seventeen percent of the patients were aged above 35 years, and 27% of the patients had a previous history of failed COS-infertility treatment cycle (s). The mean duration of infertility was 5 ± 3.8 years. The participant flow is depicted shown in Figure 2 . Demographics and baseline characteristics are provided in Table 1 . Participant flow in the study. r-hFSH = Recombinant human follicle-stimulating hormone, IUI = Intrauterine insemination, IVF = In vitro fertilization Demographics and relevant baseline characteristics SD=Standard deviation, BMI=Body mass index, COS=Controlled ovarian stimulation Four patients tested positive for anti-r-hFSH antibodies after the administration of r-hFSH at different time points; all of these subsequently turned negative [ Table 2 ]. None of these samples tested positive for neutralizing anti-r-hFSH antibodies. None of these patients developed any hypersensitivity reaction or were judged to have reduced response to r-hFSH. Immunogenicity results ADA=Antidrug antibody The clinical course and time points in the patients who seroconverted during the study are summarized in Table 3 . Clinical course in individuals with antidrug antibody ADA=Antidrug antibody, CC=Clomiphene citrate, IUIs=Intrauterine inseminations, IVF= in vitro fertilization, r-hFSH=Recombinant-human follicle-stimulating hormone, hCG=Human chorionic gonadotropin The efficacy results in the IUI and IVF cycles are summarized in Table 4 . The mean total dose and duration of r-hFSH treatment used for ovarian stimulation in IUI cycles were 816 (±543) IU and 8.1 (±3.6) days, respectively. The average number of follicles with diameter >16 mm at the end of ovarian stimulation was 1.7 (±1.6). In 202 (83.5%) of the cycles, there was development of at least 1 follicle >16 mm in diameter. Thirty (12.4%) IUI cycles showed biochemical pregnancy test (serum β-hCG test) positive on day 14 after IUI procedure, while 28 (11.6%) of these showed clinical pregnancy result in positive identified by ultrasonography (USG) at day 30 after IUI procedure. Efficacy data *Not suitable as efficacy endpoint for IVF cycles. Percentages have been calculated with number of cycles in each assigned treatment group as denominator. IUI=Intrauterine insemination, IVF= In vitro fertilization, r-hFSH=Recombinant-human follicle-stimulating hormone, SD=Standard deviation The mean total dose and duration of r-hFSH treatment used for ovarian stimulation in IVF cycles were 2183 (±869) IU and 9.5 (±2.2) days, respectively. The average number of follicles with diameter >16 mm at the end of ovarian stimulation was 5.1 (±3.5). Thirty-five (32.7%) IVF cycles showed biochemical pregnancy test (serum β-hCG test) positive at day 14 after embryo transfer (ET), while 32 (29.9%) of these showed clinical pregnancy result in positive identified by USG at day 30 after ET. A total of seven AEs were reported by six patients [ Table 5 ]. Out of seven AEs, five were graded as mild, one was moderate, and one was severe. There were three AEs, which were reported as serious since they required hospitalization of the patients. These included one case of ovarian hyperstimulation syndrome (OHSS), ectopic pregnancy, and viral fever. The patient with ectopic pregnancy required endoscopic removal of the ectopic growth, which resulted in the removal of a section of her right fallopian tube. Two AEs (both OHSS) were judged to be related to the study treatment. One of these patients developed mild OHSS with mild abdominal pain and bloating which subsided with symptomatic medication. The other patient developed moderate OHSS with moderate abdominal pain and abdominal distension and was hospitalized and administered oral fluids and intravenous colloids for a day. All the AEs were resolved. Safety data *As per MedDRA. OHSS=Ovarian hyperstimulation syndrome, MedDRA=Medical dictionary for regulatory authorities

Based on existing literature on immunogenicity of FSH (including r-hFSH), the anticipated incidence of ADA to r-hFSH was very low. The results reported here confirm the low immunogenic potential of the tested preparation of r-hFSH and provide further evidence for the efficacy and safety of this preparation. Although not routinely investigated in clinical practice, an immune response could be responsible for unexpected clinical outcomes. The clinical manifestations may include loss of efficacy, neutralization of the natural counterpart, and immune effects such as allergy, anaphylaxis, or serum sickness.[ 2 10 11 ] The development of such antibodies is usually exposure dependent, and the risk increases in line with several factors, such as time on treatment, changes in the drug formulation, alterations to the protein structure, and the manufacturing process, variations between batches of the drug, the route of administration, the dose level, and the frequency of dosing.[ 6 ] In the present study, ADA was detected in only 1.4% (4 out of 285) of the enrolled patients after the administration of r-hFSH, with none of them identified to be neutralizing in nature. Notably, the development of these ADA was transient in all patients, and further testing showed the absence of ADA in all these patients. Further assessment of the patients with detected ADA in relation to their efficacy, and safety data also did not reveal any impact of the ADA. None of these patients developed any adverse effect or local injection site reactions. Three of the four patients, who completed COS, achieved the development of mature follicle for further IUI procedure; the cycle was canceled in the fourth patient for the performance of a surgery and could not be assessed for efficacy. The present study, being a phase 4 study, did not have stringent controls on the regimen of r-hFSH used for COS, which was essentially left at the investigator's clinical judgment. The average success rate with IUI is considered to be around 13%.[ 12 ] A 2014 meta-analysis reported clinical pregnancy rates ranging from 7.6% to 22.6% per cycle in patients, who received r-hFSH for COS-IUI.[ 13 ] A retrospective study with 2019 IUI cycles from France reported a pregnancy rate of 14.8%.[ 14 ] A recent retrospective analysis of 800 IUI cycles from a tertiary center in India reported a clinical pregnancy rate of 14.1%.[ 15 ] The clinical pregnancy rate of 11.6% in IUI cycles observed in this study is in line with the literature-reported data. With respect to success rates in IVF, a 2011 Cochrane review of randomized controlled studies comparing r-hFSH versus urinary gonadotrophins in ART cycles reported clinical pregnancy rates ranging from 16.7% to 56% with r-hFSH, with most studies reporting a rate around 24%–27%.[ 16 ] The clinical pregnancy rate of about 30% with IVF in the present study is in conformity with these data. OHSS (one mild and one moderate) was reported in only two (0.7%) patients assigned to IVF treatment. There was one case of ectopic pregnancy, another complication known to occur with higher frequency in those undergoing ART. The only other serious AE reported was a case of viral fever, which resolved with symptomatic treatment and was judged to be unrelated to the treatment procedure. There were three other AEs reported, all of which were nonserious and were deemed unrelated by the investigators. We acknowledge that the fixed time points for immunogenicity assessment may fail to identify all patients who develop an antibody response, as there can be wide variations in the onset and duration of such a response; and continuing pharmacovigilance will be needed.

The results of this study support the low immunogenic potential of the r-hFSH, consistent with existing immunogenicity data for r-hFSH. The overall safety and efficacy profile for the tested preparation is also in line with that available in the literature, and no new safety concerns were identified. This study was financially supported by Bharat Serums and Vaccines Limited. Ganesh H. Divekar and James John are employees of Bharat Serums and Vaccines Limited, India.

This was a prospective, multicenter, open-label, controlled study to assess the immunogenicity of an r-hFSH preparation (Foligraf ® , manufactured by Bharat Serums and Vaccines Limited, Mumbai, India). Although the choice of gonadotropin (only r-hFSH) and the minimum and maximum dose of r-hFSH was fixed, the choice of IUI/IVF and other treatment protocols was at the investigator's discretion. The study was conducted at 12 centers (ten centers in India and two centers in Vietnam). The study protocol was approved by the Indian and Vietnamese drug regulatory authorities and the institutional ethics committees of all the participating centers. The study was registered on Clinical Trials Registry-India (CTRI/2014/08/004886). The study was performed in accordance with the principles of the Declaration of Helsinki, the International Conference on Harmonization Guidelines for Good Clinical Practice, and local regulatory requirements. All participants provided written informed consent. Premenopausal women aged 20–40 years with infertility requiring COS as a part of one, two, or three successive cycles, of either IUI or IVF, were eligible for the study. Additional main inclusion criterion was the presence of normal reproductive tract anatomy compatible with pregnancy. The main exclusion criteria were history of receiving injectable gonadotropins within the past 3 months; severe endometriosis; pelvic pathology or chronic systemic disease that would compromise pregnancy; pregnancy, lactation, or contraindication to pregnancy; history of abuse of alcohol or drugs; history of tumors of the ovary, breast, adrenal gland, pituitary, or hypothalamus and malformation of sexual organs incompatible with pregnancy; and history of hypersensitivity to any gonadotropin. A minimum of 250 patients was planned to be included in the study. The order of the study activities is depicted in Figure 1 . Order of study activities. IUI = Intrauterine insemination, IVF = In vitro fertilization, ET = Embryo transfer, USG = Ultrasonography The primary outcome measure was the incidence of development of anti-drug antibodies (ADA) and their neutralization potential. The secondary outcome measures included follicles >16 mm, total dose and duration of r-hFSH, biochemical (serum β-human chorionic gonadotropin [hCG] test) pregnancy rate, clinical pregnancy rate, cycle cancellation rate, and incidence of adverse events (AEs). The efficacy outcomes were analyzed separately for patients assigned to IUI and IVF treatments. Detection of ADA – A radioimmunoprecipitation assay was developed at Bioanalytical Laboratory, Syngene International Limited, Bengaluru, India, for the detection of ADA in human serum. The method was validated following the current regulatory guidelines[ 5 6 7 ] by the United States Food and Drug Administration and European Medicines Agency (EMA). The assay cutpoint (for binding percentage) was determined statistically from the level of binding seen with the use of negative controls during assay validation and was based on a targeted false-positive rate of 5% The assay involved the following: Screening assay – Here, the aim was to determine the presence of antibodies based on their ability to recognize the relevant antigenic determinants in the therapeutic protein.[ 8 ] It involved incubation of positive control antibodies/ADA prepared in human serum with Iodine-125 ( 125 I)-labeled r-hFSH. Precipitation of the antigen–antibody complex was accomplished by addition of polyethylene glycol (PEG) to precipitate the 125 I-labeled FSH/anti-FSH antibody complexes. The radioactivity in the bound fraction was measured using a gamma counter as counts per minute, and the results were expressed as a binding percentage. The binding percentage was directly proportional to the concentration of anti-FSH antibodies in the sample Confirmatory assay – This step aimed to minimize false-positive results from the initial screening.[ 8 ] It followed the same principle as the screening assay, but with an additional competitive step, in which the positive controls and study samples were retested following prior incubation with and without the excess drug (r-hFSH). Due to the prior incubation, the ADA was no longer freely available. These mixtures were further incubated with 125 I-labeled r-hFSH. After precipitation of the antibody complexes with PEG, the radioactivity in the bound fraction was measured using a gamma counter. As compared to positive controls, which were not incubated with the drug, positive controls and test samples incubated with the drug showed reduction in binding percentage. Test samples with values above the assay cutpoint were confirmed positive Antibody titer check – The samples, which were confirmed positive with the detection assay, were then checked for antibody titer. The reciprocal of the highest dilution of the sample with a response that remained above the assay cutpoint was considered as the antibody titer Assessment of neutralization potential of ADA – The ADA-positive samples are required to be further characterized to check if they are capable of neutralizing the biological function of the drug.[ 8 ] In many cases, the ADA raised against a drug may not neutralize the biological activity of the drug, and such ADA does not interfere with the efficacy of the drug. Hence, it is essential to evaluate the ADA-positive samples for neutralizing potential in a functional assay. The positive ADA samples were further characterized for binding of ADA to the drug receptor and to evaluate if the ADA were neutralizing or nonneutralizing. This was achieved by a cell-based radioimmunoassay, developed and validated at Chelatec SAS, France. This competitive binding assay used transfected Chinese Hamster Ovary cells expressing a receptor for FSH and 125 I-labeled r-hFSH as a tracer and was aimed to evaluate the clinical relevance of any persistent ADA response. The assay was validated for various assay parameters as per the international guidelines before testing the samples tested positive for ADA.[ 5 ] The presence of neutralizing antibodies in the test serum samples was expected to decrease the binding of the 125 I-labeled r-hFSH to the receptor, giving rise to reduction in the radioactivity. The binding and neutralization of radiolabeled drug and the neutralization assay were optimized using a polyclonal anti-FSH antibody raised in rabbits. Detection of ADA – A radioimmunoprecipitation assay was developed at Bioanalytical Laboratory, Syngene International Limited, Bengaluru, India, for the detection of ADA in human serum. The method was validated following the current regulatory guidelines[ 5 6 7 ] by the United States Food and Drug Administration and European Medicines Agency (EMA). The assay cutpoint (for binding percentage) was determined statistically from the level of binding seen with the use of negative controls during assay validation and was based on a targeted false-positive rate of 5% The assay involved the following: Screening assay – Here, the aim was to determine the presence of antibodies based on their ability to recognize the relevant antigenic determinants in the therapeutic protein.[ 8 ] It involved incubation of positive control antibodies/ADA prepared in human serum with Iodine-125 ( 125 I)-labeled r-hFSH. Precipitation of the antigen–antibody complex was accomplished by addition of polyethylene glycol (PEG) to precipitate the 125 I-labeled FSH/anti-FSH antibody complexes. The radioactivity in the bound fraction was measured using a gamma counter as counts per minute, and the results were expressed as a binding percentage. The binding percentage was directly proportional to the concentration of anti-FSH antibodies in the sample Confirmatory assay – This step aimed to minimize false-positive results from the initial screening.[ 8 ] It followed the same principle as the screening assay, but with an additional competitive step, in which the positive controls and study samples were retested following prior incubation with and without the excess drug (r-hFSH). Due to the prior incubation, the ADA was no longer freely available. These mixtures were further incubated with 125 I-labeled r-hFSH. After precipitation of the antibody complexes with PEG, the radioactivity in the bound fraction was measured using a gamma counter. As compared to positive controls, which were not incubated with the drug, positive controls and test samples incubated with the drug showed reduction in binding percentage. Test samples with values above the assay cutpoint were confirmed positive Screening assay – Here, the aim was to determine the presence of antibodies based on their ability to recognize the relevant antigenic determinants in the therapeutic protein.[ 8 ] It involved incubation of positive control antibodies/ADA prepared in human serum with Iodine-125 ( 125 I)-labeled r-hFSH. Precipitation of the antigen–antibody complex was accomplished by addition of polyethylene glycol (PEG) to precipitate the 125 I-labeled FSH/anti-FSH antibody complexes. The radioactivity in the bound fraction was measured using a gamma counter as counts per minute, and the results were expressed as a binding percentage. The binding percentage was directly proportional to the concentration of anti-FSH antibodies in the sample Confirmatory assay – This step aimed to minimize false-positive results from the initial screening.[ 8 ] It followed the same principle as the screening assay, but with an additional competitive step, in which the positive controls and study samples were retested following prior incubation with and without the excess drug (r-hFSH). Due to the prior incubation, the ADA was no longer freely available. These mixtures were further incubated with 125 I-labeled r-hFSH. After precipitation of the antibody complexes with PEG, the radioactivity in the bound fraction was measured using a gamma counter. As compared to positive controls, which were not incubated with the drug, positive controls and test samples incubated with the drug showed reduction in binding percentage. Test samples with values above the assay cutpoint were confirmed positive Antibody titer check – The samples, which were confirmed positive with the detection assay, were then checked for antibody titer. The reciprocal of the highest dilution of the sample with a response that remained above the assay cutpoint was considered as the antibody titer Assessment of neutralization potential of ADA – The ADA-positive samples are required to be further characterized to check if they are capable of neutralizing the biological function of the drug.[ 8 ] In many cases, the ADA raised against a drug may not neutralize the biological activity of the drug, and such ADA does not interfere with the efficacy of the drug. Hence, it is essential to evaluate the ADA-positive samples for neutralizing potential in a functional assay. The positive ADA samples were further characterized for binding of ADA to the drug receptor and to evaluate if the ADA were neutralizing or nonneutralizing. This was achieved by a cell-based radioimmunoassay, developed and validated at Chelatec SAS, France. This competitive binding assay used transfected Chinese Hamster Ovary cells expressing a receptor for FSH and 125 I-labeled r-hFSH as a tracer and was aimed to evaluate the clinical relevance of any persistent ADA response. The assay was validated for various assay parameters as per the international guidelines before testing the samples tested positive for ADA.[ 5 ] The presence of neutralizing antibodies in the test serum samples was expected to decrease the binding of the 125 I-labeled r-hFSH to the receptor, giving rise to reduction in the radioactivity. The binding and neutralization of radiolabeled drug and the neutralization assay were optimized using a polyclonal anti-FSH antibody raised in rabbits. According to the EMA guidelines[ 9 ] on r-hFSH, the immunogenicity of r-hFSH is low, and neutralizing antibodies have not been reported. Hence, a pragmatic sample size of 250 was planned for the study. Qualitative data are presented in the form of frequency, and percentage and quantitative data are represented in the form of mean ± standard deviation (SD) or median (interquartile range) as per the distribution of the variable, with the former used in case of data with normality and the latter in the absence of the same.