Dual-trigger improves the outcomes of in vitro fertilization cycles in older patients with diminished ovarian reserve: A retrospective cohort study.

The process of in vitro fertilization (IVF) involves hyperstimulation of ovaries with gonadotropins to mimic the natural cycle in producing mature oocytes. In the normal menstrual cycle, oocyte maturation occurs following the luteinizing hormone (LH) and smaller follicle stimulating hormone (FSH) surge that happens during mid-cycle. Traditionally, human chorionic gonadotropin (hCG) has been the trigger of choice for oocyte maturation due to its molecular and biological similarity with LH [ 1 ]. However, its half-life is much longer than LH, lasting for approximately days [ 2 ]. This contributes to the occurrence of ovarian hyperstimulation syndrome (OHSS) in high responders. Also, hCG lacks FSH activity, which plays a role in the in vitro maturation of oocytes [ 3 ]. Gonadotropin-releasing hormone (GnRH) agonists were first suggested for final oocyte maturation by Gonen et al. in 1990, as it is able to trigger endogenous release of both FSH and LH [ 4 ]. With a shorter mean duration of LH surge of about 34 hours, it is similar to the natural cycle duration of 48 hours [ 5 ], effectively reducing the incidence of OHSS in high responders [ 6 , 7 ]. However, some problems surfaced with the substitution of GnRH-agonists as trigger. Kummer et al. discovered that the risk of empty follicle syndrome was increased following isolated GnRH-agonist trigger due to a suboptimal LH surge [ 8 ]. Additionally, increased early pregnancy loss and decreased rates of ongoing pregnancy were noted by multiple studies [ 9 , 10 ]. As such, the idea of a dual trigger was developed [ 11 ]. Indeed, the hCG component of dual trigger could serve as a rescue trigger in case of poor response to GnRH-agonist, which occurs in about 2.71% of a study population [ 12 ]. Since its development, multiple investigations have shown the benefits of using a dual trigger for final oocyte maturation in normal responders [ 11 , 13 ], including an improvement in total number of retrieved oocytes, MII oocytes, rates of embryo implantation, clinical pregnancy, and live birth rates [ 14 ]. Evidence from available meta-analysis in 2018 involving four studies including 527 patients found a significantly improved clinical pregnancy rate following dual trigger [ 15 ]. However, for poor ovarian responders (PORs), the situation is less clear cut. Women with poor ovarian response are those who have a reduced number of follicles responsive to FSH, resulting in poor IVF outcomes [ 16 ]–posing a grave challenge to clinicians worldwide. Numerous criteria have been proposed for the definition of POR [ 17 ], but none was established as the international standard to define POR until the creation of the Bologna criteria in 2011 [ 18 ]. Presence of much heterogeneity lead to strong criticism about the clinical application of the Bologna criteria [ 19 ], triggering the development of the POSEIDON (Patient-Oriented Strategies Encompassing IndividualizeD Oocyte Number) criteria in 2016 [ 20 ]. The POSEIDON criteria brings forward the concept of “low prognosis”, and classifies all patients according to 3 factors: oocyte quality as demonstrated by age, oocyte quantity as represented by ovarian biomarkers of anti-Müllerian hormone (AMH) and/or antral follicle count (AFC), and finally ovarian response, dependent on the number of oocytes retrieved given a previous ovarian stimulation cycle. As such, it is able to aid the clinician in formulating a more tailored treatment plan for the individual groups. The focus of our study is POSEIDON group 4, with patients who are ≥ 35 years old, with AFC < 5 and/or AMH < 1.2 ng/mL. However, it is of note that the patients in POSEIDON group 4 do not display elevated FSH, which is a distinguishing feature of premature ovarian insufficiency. Based on the existing evidence of improved reproductive outcomes in normal responders, and the theoretical advantage of a more physiologic trigger while reducing risk of poor response to GnRH-agonist alone, we hypothesized that similar benefits can be obtained using dual-trigger in a low prognosis group like POSEIDON 4. Therefore, we attempted to investigate the feasibility of utilizing dual-trigger for final oocyte maturation in improving IVF outcomes of this population.

As shown in Fig 1 , out of 2,165 IVF cycles, 384 cycles fulfilled the POSEIDON group 4 criteria. Of the 384 cycles, 38 were not treated using a GnRH antagonist protocol, 8 were diagnosed with premature ovarian insufficiency, 19 with advanced maternal age of more than 45 years old, 3 were cancer patients who underwent chemotherapy and/or radiotherapy, and 8 had incomplete dataset. Those cycles were excluded from the study. Of the remaining 194 cycles in the dual-trigger group and 114 in the hCG-only trigger group, a further 34 cycles were excluded in the dual-trigger group, and 22 cycles excluded in the hCG-only group due to no retrieved oocytes or no viable embryos. As such, 160 frozen-thawed embryo transfer cycles from the dual-trigger group and 92 frozen-thawed embryo transfer cycles from the hCG-only group were available for analysis. IVF, in vitro fertilization; ICSI, intracytoplasmic sperm injection; GnRH, gonadotropin-releasing hormone; POI, primary ovarian insufficiency; FET, frozen embryo transfer. Comparisons between the populations of two groups revealed no difference in patient age, body mass index, infertility duration, previous IVF attempts, primary or secondary infertility, cause of infertility, basal FSH, basal luteinizing hormone (LH), basal estradiol (E2), AFC and AMH ( Table 1 ). Data are presented as mean ± standard deviation and %. IVF, in vitro fertilization; FSH, follicle-stimulating hormone; LH, luteinizing hormone; E2, estradiol Cycle characteristics between the two groups are presented in Table 2 . There were no differences in terms of stimulation duration, types of gonadotropins, total gonadotropin dose and percentage of growth hormone supplementation. Pre-ovulatory follicles > 10 mm (3.7±2.6 vs. 3.4±1.5, p = 0.573) and pre-ovulatory follicles > 14 mm (2.9±2.2 vs. 2.7±0.9, p = 0.191) on trigger day were similar between the two groups. However, the number of retrieved oocytes (3.3±2.7 vs. 1.6±1.5, p <0.001), metaphase II oocytes (2.6±2.0 vs. 1.3±1.0, p <0.001), fertilized oocytes (2.4±2.1 vs. 1.2±1.0, p <0.001), day-3 embryos (2.2±1.9 vs. 1.2±1.0, p < 0.001) and top-quality day-3 embryos (0.9±1.3 vs. 0.2±0.5, p <0.001) were significantly higher in the dual-trigger group compared with the hCG-only group. Moreover, methods of fertilization, fertilization rate and cancellation rate were similar between the two groups. Data are presented as mean ± standard deviation and %. rFSH, recombinant follicle-stimulating hormone; rLH, recombinant luteinizing hormone; HMG, human menopausal gonadotropin; GH, growth hormone; IVF, in vitro fertilization; ICSI, intracytoplasmic sperm injection Pregnancy outcomes between the two groups are presented in Table 3 . Endometrium thickness was similar between the two groups. However, the number of transferred embryos (2.1±1.0 vs. 1.4±0.8, p <0.001) and percentage of at least one top-quality embryos transfer (62.5% vs. 23.9%, p <0.001) were significantly higher in the dual-trigger group than in the hCG-only group. Of note, the dual-trigger group performed superiorly in terms of implantation rate (14.4±30.0% vs. 5.4±18.8%, p = 0.004), clinical pregnancy rate (23.1% vs. 8.7%, p = 0.004) and live birth rate (17.5% vs. 5.4%, p = 0.006). Furthermore, there was no significant difference in miscarriage rate between the two groups. Data are presented as mean ± standard deviation and %. GEEs were performed to determine whether dual-trigger use had a beneficial effect on clinical pregnancy rate and live birth rate in Table 4 . Confounding variables such as age, BMI, duration of infertility, previous IVF attempts, basal FSH, AFC, AMH, number of pre-ovulatory follicles > 10 mm on trigger day, and number of pre-ovulatory follicles > 14 mm on trigger day were included in the analysis. The analysis showed that dual-trigger use was positively associated with live birth rate (OR = 3.16, 95% CI 1.06–9.38, p = 0.039) and clinical pregnancy rate (OR = 4.30, 95% CI 1.38–13.43, p = 0.012). Furthermore, AMH was shown to be a positive independent factor affecting clinical pregnancy rate (OR = 7.20, 95% CI 1.31–39.61, p = 0.023). OR, odds ratio; CI, confidence interval; BMI, body mass index; IVF, in vitro fertilization; FSH, follicle- stimulating hormone; AFC, antral follicle count; AMH, anti-Müllerian hormone.

Our data suggest that dual trigger might increase both oocyte and embryo yields, as well as clinical pregnancy rates and live birth rates in patients fulfilling the POSEIDON group 4 criteria. However, large-scale randomized controlled studies are needed to confirm these findings.

The retrospective cohort study was performed at the reproductive medical center of Kaohsiung Veterans General Hospital, in Kaohsiung, Taiwan from January 2012 through December 2017. The study protocol was approved by the institutional review board at Kaohsiung Veterans General Hospital, with the identifier VGHKS19-CT8-05, and conforms to the “Declaration of Helsinki for Medical Research involving Human Subjects.” The need for consent was waived by the ethics committee due to its retrospective design. Data was compiled from electronic medical records and cycle flow sheets for patients within the study period. Patients who underwent IVF cycles and fulfilled the criteria for POSEIDON group 4 (age ≥ 35 years, with AFC < 5 and/or AMH < 1.2 ng/mL) were included in this study. IVF is the treatment of choice for this group of patients characterized by advanced maternal age and low ovarian reserve due to its relatively higher success rate compared to other treatment modalities. A total of 384 cycles met the criteria during the study period. Exclusion criteria were as follows: (i) Patients who did not receive GnRH antagonist protocol, (ii) patients with premature ovarian insufficiency, (iii) patients over the age of 45, (iv) cancer patients who have received chemotherapy or radiotherapy, and (v) patients with incomplete data. Following application of the exclusion criteria, a total of 308 cycles were included for study and then divided into dual-trigger and hCG trigger groups. The choice of hCG alone or dual-trigger depended on the physician, with cycles before 2015 being mostly hCG-only trigger and those after 2015 being mostly dual-trigger. In the dual-trigger group (n = 194), patients received final oocyte maturation with GnRH-agonist and hCG. In the hCG trigger group (n = 114), patients received only hCG for final oocyte maturation. Only patients receiving GnRH antagonist protocol were included in this study. Following a baseline hormone screen and transvaginal ultrasound for antral follicles, controlled ovarian stimulation was initiated within 5 days of the menstrual cycle, with recombinant follicle stimulating hormone (rFSH, Gonal-F, Merck Serono S.p.A., Modugno, Italy), rFSH plus recombinant luteinizing hormone (Pergoveris, Merck Serono SA, Aubonne, Switzerland), human menopausal gonadotropin (Merional, IBSA Institut Biochimique S.A., Lamone, Switzerland), or corifollitropin alfa (Elonva, Vetter Pharma-Fertigung GmbH & Co, KG, Ravensburg, Germany). Furthermore, addition of growth hormone (Saizen, Merck Serono SA, Aubonne, Switzerland) was given during controlled ovarian stimulation at physician’s discretion. Patient response was monitored during the IVF cycle with serial transvaginal ultrasound for follicular measurements and hormone profiles. Dosage was adjusted depending on follicular response and previous response to gonadotropins. Daily GnRH antagonist injections (Cetrotide 0.25mg, Pierre Fabre Medicament Production, Aquitaine Pharm International, Idron, France or Orgalutran 0.25mg, Vetter Pharma-Fertigung GmbH & Co, KG, Ravensburg, Germany) were administered when the leading follicle reaches 12–14 mm in diameter, up till the date of final oocyte maturation. Patient then either received dual-trigger with combined recombinant hCG (Ovidrel 250μg, Merck Serono S.p.A., Modugno, Italy) with GnRH agonist (Lupro 2mg, Nang Kuang Pharmaceutical Co, Ltd., Tainan, Taiwan) or recombinant hCG alone. Trigger is given when at least one leading follicle reaches a mean diameter of 18 mm. Transvaginal ultrasound-guided oocyte retrieval was performed 36 h following administration of trigger. Whether fertilization was conducted by conventional IVF or intracytoplasmic sperm injection (ICSI) depended on semen analysis results or prior fertilization condition. Embryos were evaluated and graded according to the criteria established by the Istanbul consensus workshop [ 21 ]. The number of blastomeres, the percentage of fragmentation and the variation in blastomere symmetry were assessed. Top-quality cleavage stage embryos were determined as those with the following features: six cells or greater on day 3, less than 10% fragmentation and symmetric blastomeres. All embryos were cryopreserved by vitrification under a freeze-all strategy on the third day after oocyte retrieval for subsequent frozen embryo transfer. As such, no embryo qualified for preimplantation genetic testing for aneuploidies (PGT-A). Vitrification was carried out via a two-step exposure to equilibrium and vitrification solutions. The embryo is first exposed to the equilibrium solution for 15 minutes, followed by exposure to the vitrification solution for 1 minute and 30 seconds. The embryo is then loaded onto the propylene strip of Cryotop (Kitazato, Bio Pharma Co., Tokyo, Japan) in an open system with minimal solution, and then rapidly plunged into liquid nitrogen at -196°C. An artificial cycle was used for endometrial preparation. Endometrium was prepared with daily oral estradiol (Ediol 8mg, Synmosa Biopharma Corporation, Hsinchu County, Taiwan) and estradiol gel (Oestrogel gel, Besins, Drogenbos, Belgium) beginning before cycle day 5. When the endometrial thickness reaches at least 8mm, daily progesterone, including intravaginal gel (Crinone 8% gel, Merck Serono, Hertfordshire, UK) and oral dydrogesterone (Duphaston 40mg, Abbott, Olst, the Netherlands), were given simultaneously as luteal phase support. Additionally, aspirin (ASPIRIN PROTECT 100mg, Bayer AG, Leverkusen, Germany) was also prescribed routinely. Following assisted-hatching via laser zona drilling, the frozen-thawed embryos were transferred under transabdominal sonographic assistance. Progesterone supplementation was administered until 8–10 gestational weeks upon confirmation of pregnancy. The primary outcome measure was live birth rate, defined as the delivery of a viable fetus beyond 24 weeks of gestation. Secondary outcome measures included the number of retrieved oocytes, number of mature oocytes, number of fertilized oocytes, number of day-3 embryos, number of top-quality day-3 embryos, implantation rate (calculated from the number of gestation sacs with fetal heart seen on ultrasound scan divided by the total number of transferred embryos), cancellation rate, miscarriage rate, and clinical pregnancy rate. Cancellation rate was defined as those with no retrieved oocytes, or no viable embryos, while miscarriage rate refers to pregnancy loss before 24 weeks of gestation, and clinical pregnancy rate was defined by the presence of a fetal heartbeat at 6–7 weeks of a pregnancy. Kolmogovor-Smirnov test was used to evaluate normality of quantitative variables. Quantitative variables were evaluated using the independent t-test. Chi-Square test was used to evaluate categorical variables. Odds ratios (ORs) and 95% confidence intervals (CIs) of live birth and clinical pregnancy were assessed using generalized estimating equations (GEEs), after adjusting for confounders. We adopted GEEs to account for correlations between multiple cycles from the same patient. Key factors including age, BMI, infertility duration, previous IVF attempts, basal FSH, AFC, AMH, number of pre-ovulatory follicles > 10 mm on trigger day, and number of pre-ovulatory follicles > 14 mm on trigger day, were identified as confounders for analyses. Analyses were performed using the Statistical Package for Social Sciences (SPSS) version 20.0 (Chicago, IL, USA). P < 0.05 was considered statistically significant.

(XLS) Click here for additional data file.