https://orcid.org/0000-0002-6264-3787
Abstract
What are outcome and procedural differences when using the semi-automated closed Gavi® device versus the manual open Cryotop® method for vitrification of pronuclear (2PN) stage oocytes within an IVF program?
A semi-automated closed vitrification method gives similar clinical results as compared to an exclusively manual, open system but higher procedure duration and less staff convenience.
A semi-automated closed vitrification device has been introduced to the market, however, little evaluation of its performance in a clinical setting has been conducted so far.
This prospective, randomised, open non-inferiority trial was conducted at three German IVF centers (10/2017–12/2018). Randomization was performed on day of fertilization check, stratified by center and by indication for vitrification (surplus 2PN oocytes in the context of a fresh embryo transfer (ET) cycle or ‘freeze-all’ of 2PN oocytes).
The study population included subfertile women, aged 18–40 years, undergoing IVF or ICSI treatment after ovarian stimulation, with 2PN oocytes available for vitrification. The primary outcome was survival rate of 2PN oocytes at first warming procedure in a subsequent cycle and non-inferiority of 2PN survival was to be declared if the lower bound 95% CI of the mean difference in survival rate excluded a difference larger than 9.5%; secondary, descriptive outcomes included embryo development, pregnancy and live birth rate, procedure time and staff convenience.
The randomised patient population consisted of 149 patients, and the per-protocol population (patients with warming of 2PN oocytes for culture and planned ET) was 118 patients. The survival rate was 94.0% (±13.5) and 96.7% (±9.7) in the Gavi® and the Cryotop® group (weighted mean difference −1.6%, 95% CI −4.7 to 1.4, P = 0.28), respectively, indicating non-inferiority of the Gavi® vitrification/warming method for the primary outcome. Embryo development and the proportion of top-quality embryos was similar in the two groups, as were the pregnancy and live birth rate. Mean total procedure duration (vitrification and warming) was higher in the Gavi® group (81 ± 39 min vs 47 ± 15 min, mean difference 34 min, 95% CI 19 to 48). Staff convenience assessed by eight operators in a questionnaire was lower for the Gavi® system. The majority of respondents preferred the Cryotop® method because of practicality issues.
The study was performed in centers with long experience of manual vitrification, and the relative performance of the Gavi® system as well as the staff convenience may be higher in settings with less experience in the manual procedure. Financial costs of the two procedures were not measured along the trial.
With increasing requirements for standardization of procedures and tissue safety, a semi-automated closed vitrification method may constitute a suitable alternative technology to the established manual open vitrification method given the equivalent clinical outcomes demonstrated herein.
The trial received no direct financial funding. The Gavi® instrument, Gavi® consumables and staff training were provided for free by the distributor (Merck, Darmstadt, Germany) during the study period. The manufacturer of the Gavi® instrument had no influence on study protocol, study conduct, data analysis, data interpretation or manuscript writing. J.H. has received honoraria and/or non-financial support from Ferring, Merck and Origio. G.G. has received honoraria and/or non-financial support from Abbott, Ferring, Finox, Gedeon Richter, Guerbet, Merck, MSD, ObsEva, PregLem, ReprodWissen GmbH and Theramex. The remaining authors have no competing interests.
ClinicalTrials.gov NCT03287479.
19 September 2017.
10 October 2017.
Introduction
Cryopreservation of human gametes and embryos has become an essential practice in assisted reproduction technology. It is used for various indications such as storage of surplus embryos, fertility preservation, deferring embryo transfer (ET) to prevent ovarian hyperstimulation syndrome (Devroey et al., 2011; Griesinger et al., 2011; Herrero et al., 2011), postponing ET to a subsequent cycle with purportedly better uterine receptivity (Shapiro et al., 2011; Roque et al., 2013), avoiding the negative impact of prematurely elevated progesterone levels (Shapiro et al., 2010; Griesinger et al., 2013; Hill et al., 2015) or to allow complex genetic testing requiring extended periods of time.
The last decade has seen a shift from conventional slow freezing toward vitrification since this technique achieves higher oocyte and embryo survival rates when compared to conventional slow freezing (Rienzi et al., 2017). Therefore, vitrification is being increasingly applied in ART clinics worldwide and has been recognised as the method of choice for cryopreservation.
Several vitrification protocols have been described and a variety of commercial kits are available. They differ from one another in the composition of media (the type and concentration of cryoprotectants), equilibration time and temperature, the carrier devices and the cooling, storage and warming methods, as well as in the volume of vitrification solution surrounding the embryo.
To date, the gold standard is arguably the Cryotop® method. Cryotop® is an open vitrification device and its design offers extremely high cooling and warming rates. Several studies have demonstrated that this system achieves oocyte and embryo survival rates of >90% (Kuwayama et al., 2005; Loutradi et al., 2008; Godsen, 2011).
However, manual vitrification is one of the most time-consuming and labor-intensive procedures in an IVF laboratory. Its success depends on operator skills and dexterity as it requires embryologists to transfer single oocytes or embryos from one solution to another and to handle minimal volumes and small appliances. Hence, it is impossible to control and standardise the many variables potentially affecting outcomes, such as actual temperatures, exact incubation times and diffusion gradients of solutions at the embryo level. Although some IVF clinics employ embryologists solely dedicated to this task, success rates may vary between clinics and even within the operating staff of one center.
Another limitation of the open vitrification method is that oocytes or embryos come into direct contact with liquid nitrogen which involves the potential exposure to pathogens and, as a consequence, the potential risk of cross-contamination during storage (Bielanski et al., 2003; Vajta et al., 2015). Nevertheless, the use of open vitrification systems offers higher cooling and warming rates when compared to closed systems (Roy et al., 2014; Vajta et al., 2015) and open systems are therefore usually favored from an outcome perspective.
To standardise the vitrification procedure among operating staff and clinics, as well as to reduce labor intensity and to eliminate the potential risk of cross-contamination, Gavi® (Genea Biomedx Pty Ltd, Sydney, Australia), a semi-automated closed vitrification system, was developed.
The few, so far available studies comparing the efficacy of the semi-automated closed Gavi® system and the manual Cryotop® method report comparable results with regard to survival rate, embryo development, pregnancy rate and live birth rate (Roy et al., 2014; Hobson et al., 2016; Miwa et al., 2020). Recently, the first pregnancies following transfer of blastocysts vitrified with Gavi® (Dal Canto et al., 2019) and the first live birth after the utilization of Gavi® vitrified-warmed oocytes (Brunetti et al., 2021) have been reported. However, to date, the published studies and case reports are of limited validity due to small sample size and retrospective analysis of data. The aim of our study was to compare the vitrification utilizing the Gavi® system versus a conventional vitrification method in a randomised controlled trial with the primary outcome being post-warming survival rate in human IVF.
Materials and methods
Study design
This study was a multi-centric (n = 3), prospective, randomised, investigator-initiated, open non-inferiority trial conducted in Germany between 10/2017 and 11/2020. Institutional review board approval was obtained (Ethical Review Board of the University of Luebeck, AZ17-093, 12 May 2017), and informed consent was given by all patients included in this study. The study protocol was prospectively registered at clinical trials.gov (NCT03287479). Operating staff were provided with adequate training in the Gavi® vitrification and warming procedure by the company’s product specialists prior to study initiation. The participation in this study was limited to one vitrified-warmed ET cycle per patient with follow up of pregnant patients to live birth.
Population and IVF/ICSI treatment
Women aged 18–40 years with no history of low response in previous stimulated cycles and with no evidence of uterine pathology undergoing IVF or ICSI after ovarian stimulation were considered eligible and screened for inclusion on day of oocyte retrieval. Preimplantation genetic testing cycles were excluded. There were no restrictions on the ovarian stimulation protocol, FSH dosage nor on type or dose of drug for triggering of final oocyte maturation. Oocyte retrieval was performed by transvaginal ultrasound-guided follicle aspiration 36 ± 2 h after the ovulation trigger. IVF and ICSI procedures were performed according to the standard operating procedures of the investigator site. The fertilization check was performed 17 ± 2 h post-insemination on Day 1 after oocyte retrieval.
Randomization
Study participants were checked for eligibility during ovarian stimulation and written informed consent was received at the latest on day of oocyte retrieval. On the morning after oocyte retrieval, the fertilization check was performed and cases presenting ≥ 2 surplus pronuclear (PN) stage oocytes (surplus group) and cases of elective cryopreservation of all PN stage embryos (freeze-all group) were randomised to either the Cryotop® vitrification/warming procedure (control group) or the Gavi® vitrification/warming procedure (study group). Randomization was performed between 10/2017 and 12/2018 and was stratified by center and by indication for vitrification (surplus or freeze-all). Randomization was performed by a laboratory staff member by a computer-generated allocation list in a ratio of 1:1 in blocks of four. Allocation concealment was achieved by opaque and sequentially numbered envelopes (per center). Each patient could be randomised only once.
Cryotop® vitrification procedure
For vitrification, the Kitazato vitrification kit VT601 (Gynemed, Lensahn, Germany), containing the permeable cryoprotectants ethylene glycol (EG) and dimethyl sulfoxide (DMSO), trehalose acting as extracellular cryoprotectant and supplemented hydroxypropyl cellulose (HPC), was used. The vitrification procedure was performed according to the manufacturer’s instructions at room temperature. Initially, a maximum of six PN stage oocytes were placed in a center well dish containing 600 μl of equilibration solution (ES) for 12–15 min (until initial size was regained after shrinkage). After equilibration, they were transferred to a center well dish containing 600 μl of vitrification solution (VS) and were flushed three times at three different spots within the dish to displace ES surrounding the zygotes. Subsequently, a maximum of two PN stage oocytes were loaded onto the Cryotop®s transparent tip in a volume of less than 0.1 μl of VS. The Cryotop® (Gynemed, Lensahn, Germany) was then quickly immersed into liquid nitrogen vertically and gently stirred for a few seconds and finally plunged completely into liquid nitrogen. Steps were repeated for remaining zygotes to be vitrified. The time span from transferring the PN stage oocytes in VS to immersion into LN2 was kept to a maximum of 60 s. Finally, the Cryotop®s were inserted into the straw caps, twisted together tightly and then transferred to a standard liquid nitrogen tank for storage.
Gavi® vitrification procedure
Following a warm-up phase after the Gavi® instrument (Merck, Darmstadt, Germany) had been switched on, the appropriate protocol for vitrification of zygotes was selected. While the Gavi® was warming up, the operating tray was loaded with tip and seal cartridges and medium cartridges containing vitrification solutions VS1 and VS2 (EG, DMSO, trehalose and supplemented human serum albumin) and subsequently was placed into the Gavi® operating tray dock. Liquid nitrogen was added to the Gavi® liquid nitrogen bucket. PN stage oocytes were first incubated in Gems VitBase (Genea Biomedx; Merck, Germany), a HEPES-buffered basic medium, at 37°C for 5 min. In the meantime, a Gavi® cassette was loaded with an appropriate number of Gavi® pods and they were prefilled with 2 µl of Gems VitBase at room temperature. Following incubation, a maximum of two PN stage oocytes were loaded into the pods, the cassette was loaded into Gavi® instrument and the program was started. On completion of the program of 14.5 min, which includes incubation in vitrification solutions and automatic sealing of pods, the cassette was manually removed, dunked in liquid nitrogen with vertical stirring and finally completely plunged into liquid nitrogen. Steps were repeated in case of remaining zygotes to be vitrified. For storage, the cassette was then transferred to a standard liquid nitrogen tank.
Embryo transfer
Depending on the cryo ET policy of the investigator site, embryos were replaced after warming of 2PN oocytes in a subsequent natural, modified natural or programmed cycle. The number of PN stage oocytes to be warmed, duration of embryo culture and number of embryos to be transferred was decided by the investigator on a case-by-case basis.
Cryotop® warming procedure
For warming, the Kitazato warming kit VT602 (Gynemed, Lensahn, Germany) was used. First, the straw cap was removed and the Cryotop®’s transparent tip was immersed in 37°C thawing solution (TS). After 1 min in TS, released zygotes were then transferred to diluent solution (DS) and incubated for 3 minutes at room temperature. Zygotes were then transferred to a culture dish prepared the day before, washed in culture medium and moved in microdroplets of fresh medium for individual continuous culture. Steps were repeated for all Cryotop®s to be warmed.
Gavi® warming procedure
For warming, the Gems warming kit (Genea Biomedx; Merck, Darmstadt, Germany) was used. A Gavi® pod was removed from the cassette (while held in liquid nitrogen) and dunked in a 37°C water bath for 3 seconds. Excess water was wiped off with a lint-free lab wipe and the Gavi® pod was placed under the microscope. The foil seal was manually removed and 10 µl of warming Solution 1 (room temperature) was immediately added directly into the Pod divot. The zygotes were to remain in warming Solution 1 within the pod for 1 min. During this time, the microscope focal plane was changed and zygotes were located. PN stage embryos were then transferred into warming Solution 2 of the previously prepared warming dish (4 well dish). After 3 min, they were transferred into warming Solution 3 for 5 min. Then zygotes were transferred in fresh warming Solution 3 for another minute. Subsequently, zygotes were transferred to the embryo culture dish prepared the day before, and washed in culture medium before being placed in individual microdroplets for continued culture. Steps were repeated for remaining Gavi® pods to be warmed.
Outcomes
The primary outcome is the survival rate of vitrified-warmed PN stage oocytes, specified as the number of vital 2PN oocytes divided by the total number of 2PN oocytes warmed per patient. Survival was evaluated 2 h post-warming and was confirmed for zygotes presenting an intact cell membrane and regular cytoplasm (Supplementary Fig. S1) and/or if cleavage was observed on day two of in vitro culture. The pre-specified minimal follow up was 6 months per patient for the primary outcome, so assessment of the primary outcome was stopped 6 months after randomization of the last patient, which was in 6/2019. The first secondary outcomes was the number of top-quality embryos on day of ET per number of warmed 2PN oocytes. An embryo was defined top-quality when displaying characteristics of optimal stage-specific development according to the Alpha/ESHRE consensus on embryo assessment (Alpha Scientists in Reproductive Medicine and ESHRE Special Interest Group of Embryology, 2011). Other secondary outcomes were: clinical pregnancy rate (fetal sac with heart beat); ongoing pregnancy rate at 10 weeks gestation; live birth rate (live born infant); procedure duration (timespan from start to end of procedure in minutes including time for paperwork and preparation and excluding the Gavi® device automated run-time of 14.5 min) adjusted for number of 2PN oocytes vitrified and warmed; and staff convenience (questionnaire completed after at least three Gavi® and Cryotop® procedures had been performed per staff member).
Sample size
Based on previous data (Golakov et al., 2018), the survival rate of 2PN oocytes after vitrification/warming using an open manual vitrification system is 95% (standard deviation 16). It was hypothesised that the semi-automated closed vitrification system is non-inferior to the manual open system in terms of cryo-survival of 2PN oocytes. To demonstrate non-inferiority of the Gavi® system at a one-sided non-inferiority margin of minus 9.5% (i.e. maximum allowed difference), group sample sizes of 61 and 61 were required (beta 0.1, alpha 0.025) using a t-test and assuming standard deviations of 16. To account for attrition (e.g. patients not utilizing vitrified PN stage oocytes because of pregnancy in a fresh cycle and/or patients not starting a vitrified-warmed ET cycle for other reasons within the study period), it was planned to randomise 180 patients in total.
Statistics
Descriptive statistics were performed by analyzing mean, standard deviation and proportions as appropriate. The differences between arms are summarised as absolute differences or relative risks with 95% CIs, as appropriate. The primary analysis uses a two-sided 95% CI of the mean difference in 2PN survival rate (weighted by center with inverse-variance, fixed effects model) with a non-inferiority margin of 9.5%, where the Gavi® system was to be declared non-inferior for this outcome if the lower bound 95% CI excluded a difference larger than 9.5%. Pregnancy and live birth rates were analyzed and adjusted by center with a fixed-effects Mantel–Haenszel model and P-values were calculated by a one-sided Fishers exact test. Parametric and non-parametric tests were used for normally distributed and non-normally distributed variables, respectively. No correction for multiplicity of testing was performed and P-values are descriptive except for the primary outcome.
Results
A total of 149 patients were randomised between 10/2017 and 12/2018 and allocated to either the Gavi® vitrification/warming method (n = 75; study group) or the Cryotop® vitrification/warming method (n = 74; control group). The patient flow through the trial is displayed in Fig. 1. In the randomised study population, a higher percentage of patients than expected (58%) were allocated to the elective vitrification group. Thus, recruitment into the trial was stopped prematurely by the principal investigator (GG) resulting in a total of 118 patients starting a vitrified-warmed ET cycle by 06/2019. At the time of stopping further recruitment and randomization, neither the total number of patients with warming nor the complete outcome data were available to the investigators. Consequently, the per-protocol population consists of 57 cases in the Gavi® group and 61 cases in the Cryotop® group. Patient demographics are presented in Table I and were comparable between the two groups as no significant differences for any demographic variables were noted.
CONSORT flow chart of patient flow.
Demographics and pre-randomization treatment characteristics (ITT population).
| Gavi® n = 75 | Cryotop® n = 74 | |
|---|---|---|
| Female age (years) | 33.2 ± 3.9 | 33.7 ± 4.0 |
| Pre-treatment AMH (ng/ml)a | 5.4 ± 4.81 | 4.2 ± 3.2 |
| At least one previous clinical pregnancy (%) | 40 (53.3) | 44 (59.5) |
| At least one previous live birth (%) | 25 (33.3) | 23 (31.1) |
| Previous IVF attempts (n) | 1.1 ± 1.2 | 1.1 ± 1.3 |
| Antagonist protocol (%) | 75 (100) | 74 (100) |
| Stimulation duration (days) | 10.4 ± 1.7 | 10.5 ± 1.5 |
| Total gonadotropin amount (IU) | 1951.5 ± 647.42 | 2099.6 ± 703.73 |
| hCG trigger (%) | 45 (60.0) | 48 (64.9) |
| GnRH-agonist trigger (%) | 30 (40.0) | 26 (35.1) |
| ICSI cases (%) | 59 (78.7) | 59 (79.7) |
| Number of COCs | 17.6 ± 8.4 | 17.0 ± 6.7 |
| Number of 2PN | 9.6 ± 5.0 | 9.2 ± 4.5 |
| Gavi® n = 75 | Cryotop® n = 74 | |
|---|---|---|
| Female age (years) | 33.2 ± 3.9 | 33.7 ± 4.0 |
| Pre-treatment AMH (ng/ml)a | 5.4 ± 4.81 | 4.2 ± 3.2 |
| At least one previous clinical pregnancy (%) | 40 (53.3) | 44 (59.5) |
| At least one previous live birth (%) | 25 (33.3) | 23 (31.1) |
| Previous IVF attempts (n) | 1.1 ± 1.2 | 1.1 ± 1.3 |
| Antagonist protocol (%) | 75 (100) | 74 (100) |
| Stimulation duration (days) | 10.4 ± 1.7 | 10.5 ± 1.5 |
| Total gonadotropin amount (IU) | 1951.5 ± 647.42 | 2099.6 ± 703.73 |
| hCG trigger (%) | 45 (60.0) | 48 (64.9) |
| GnRH-agonist trigger (%) | 30 (40.0) | 26 (35.1) |
| ICSI cases (%) | 59 (78.7) | 59 (79.7) |
| Number of COCs | 17.6 ± 8.4 | 17.0 ± 6.7 |
| Number of 2PN | 9.6 ± 5.0 | 9.2 ± 4.5 |
Depicted are mean and standard deviations or numbers and proportions.
Latest value available preceding the inclusion into the trial, measured on any day of the menstrual cycle.
n = 73/75 (2 values missing).
n = 70/75 (5 values missing because follitropin delta was used).
n = 68/74 (6 values missing because follitropin delta was used).
AMH, Anti-Muellerian Hormone; IU, International Unit; hCG, human Chorionic Gonadotropin; GnRH, Gonadotropin-Releasing-Hormone; COCs, cumulus-oocyte-complexes; 2PN, pronuclear stage oocytes.
Demographics and pre-randomization treatment characteristics (ITT population).
| Gavi® n = 75 | Cryotop® n = 74 | |
|---|---|---|
| Female age (years) | 33.2 ± 3.9 | 33.7 ± 4.0 |
| Pre-treatment AMH (ng/ml)a | 5.4 ± 4.81 | 4.2 ± 3.2 |
| At least one previous clinical pregnancy (%) | 40 (53.3) | 44 (59.5) |
| At least one previous live birth (%) | 25 (33.3) | 23 (31.1) |
| Previous IVF attempts (n) | 1.1 ± 1.2 | 1.1 ± 1.3 |
| Antagonist protocol (%) | 75 (100) | 74 (100) |
| Stimulation duration (days) | 10.4 ± 1.7 | 10.5 ± 1.5 |
| Total gonadotropin amount (IU) | 1951.5 ± 647.42 | 2099.6 ± 703.73 |
| hCG trigger (%) | 45 (60.0) | 48 (64.9) |
| GnRH-agonist trigger (%) | 30 (40.0) | 26 (35.1) |
| ICSI cases (%) | 59 (78.7) | 59 (79.7) |
| Number of COCs | 17.6 ± 8.4 | 17.0 ± 6.7 |
| Number of 2PN | 9.6 ± 5.0 | 9.2 ± 4.5 |
| Gavi® n = 75 | Cryotop® n = 74 | |
|---|---|---|
| Female age (years) | 33.2 ± 3.9 | 33.7 ± 4.0 |
| Pre-treatment AMH (ng/ml)a | 5.4 ± 4.81 | 4.2 ± 3.2 |
| At least one previous clinical pregnancy (%) | 40 (53.3) | 44 (59.5) |
| At least one previous live birth (%) | 25 (33.3) | 23 (31.1) |
| Previous IVF attempts (n) | 1.1 ± 1.2 | 1.1 ± 1.3 |
| Antagonist protocol (%) | 75 (100) | 74 (100) |
| Stimulation duration (days) | 10.4 ± 1.7 | 10.5 ± 1.5 |
| Total gonadotropin amount (IU) | 1951.5 ± 647.42 | 2099.6 ± 703.73 |
| hCG trigger (%) | 45 (60.0) | 48 (64.9) |
| GnRH-agonist trigger (%) | 30 (40.0) | 26 (35.1) |
| ICSI cases (%) | 59 (78.7) | 59 (79.7) |
| Number of COCs | 17.6 ± 8.4 | 17.0 ± 6.7 |
| Number of 2PN | 9.6 ± 5.0 | 9.2 ± 4.5 |
Depicted are mean and standard deviations or numbers and proportions.
Latest value available preceding the inclusion into the trial, measured on any day of the menstrual cycle.
n = 73/75 (2 values missing).
n = 70/75 (5 values missing because follitropin delta was used).
n = 68/74 (6 values missing because follitropin delta was used).
AMH, Anti-Muellerian Hormone; IU, International Unit; hCG, human Chorionic Gonadotropin; GnRH, Gonadotropin-Releasing-Hormone; COCs, cumulus-oocyte-complexes; 2PN, pronuclear stage oocytes.
Post-warming survival rate
In the Gavi® and in the Cryotop® group, the warming procedure was performed with an overall mean number of four to five PN stage oocytes (Table II). The post-warming survival rate was similar between the two groups: 94.0% (±13.5) and 96.7% (±9.7) in the Gavi® and the Cryotop® group. Thus, the survival rate of 2PN oocytes processed with the Gavi® system is likely to be non-inferior to the survival rate of manually vitrified PN stage oocytes (weighted mean difference −1.6%; 95% CI −4.7 to 1.4%) (Table II). A 100% survival rate was observed in 77.2% (44/57) of cases utilizing the Gavi method and in 86.9% (53/61) of cases when vitrification and warming was performed with the Cryotop® method. In neither of the two groups, was a 0% survival rate observed.
Survival rates and laboratory outcomes (PP population).
| Gavi®n = 57 | Cryotop®n = 61 | DifferenceGavi®/Cryotop®(95% CI) | P-value | |
|---|---|---|---|---|
| 2PN vitrified (elective) | 9.4 ± 6.1 | 8.8 ± 5.0 | 0.6 (−1.8, 3.0) | 0.771 |
| 2PN vitrified (surplus) | 4.6 ± 2.5 | 4.6 ± 3.2 | 0.0 (−1.4, 1.5) | 0.675 |
| 2PN warmed (total) | 4.7 ± 2.3 | 4.4 ± 1.8 | 0.3 (−0.5, 1.1) | 0.428 |
| Survival rate (%) | 94.0 ± 13.5 | 96.7 ± 9.7 | −1.6 (−4.7, 1.4) | 0.170 |
| 100% survival (n, %)a | 44/57 (77.2) | 53/61 (86.9) | −8.5 (−22.1, 5.1) | 0.220 |
| 0% survival (n, %) | 0/57 (0.0) | 0/61 (0.0) | 0 | NA |
| Embryo transfer (n, %) | ||||
| No transfer | 3/57 (5.3) | 1/61 (1.6) | 3.7 (−3.0, 10.2) | 0.277 |
| d2/d3 | 20/57 (35.1) | 23/61 (37.7) | −2.6 (−20.0, 14.7) | 0.768 |
| d5 | 34/57 (59.6) | 37/61 (60.7) | −1.1 (−18.7, 16.7) | 0.911 |
| Embryos transferred | 1.6 ± 0.6 | 1.6 ± 0.5 | 0.0 (−0.2, 0.2) | 0.818 |
| TQ embryos transferred | 0.3 ± 0.6 | 0.4 ± 0.6 | −0.1 (−0.3, 0.2) | 0.503 |
| TQ quality embryo (%) | 12.5 ± 27.6 | 10.2 ± 18.7 | 2.3 (−6.3, 10.8) | 0.584 |
| Gavi®n = 57 | Cryotop®n = 61 | DifferenceGavi®/Cryotop®(95% CI) | P-value | |
|---|---|---|---|---|
| 2PN vitrified (elective) | 9.4 ± 6.1 | 8.8 ± 5.0 | 0.6 (−1.8, 3.0) | 0.771 |
| 2PN vitrified (surplus) | 4.6 ± 2.5 | 4.6 ± 3.2 | 0.0 (−1.4, 1.5) | 0.675 |
| 2PN warmed (total) | 4.7 ± 2.3 | 4.4 ± 1.8 | 0.3 (−0.5, 1.1) | 0.428 |
| Survival rate (%) | 94.0 ± 13.5 | 96.7 ± 9.7 | −1.6 (−4.7, 1.4) | 0.170 |
| 100% survival (n, %)a | 44/57 (77.2) | 53/61 (86.9) | −8.5 (−22.1, 5.1) | 0.220 |
| 0% survival (n, %) | 0/57 (0.0) | 0/61 (0.0) | 0 | NA |
| Embryo transfer (n, %) | ||||
| No transfer | 3/57 (5.3) | 1/61 (1.6) | 3.7 (−3.0, 10.2) | 0.277 |
| d2/d3 | 20/57 (35.1) | 23/61 (37.7) | −2.6 (−20.0, 14.7) | 0.768 |
| d5 | 34/57 (59.6) | 37/61 (60.7) | −1.1 (−18.7, 16.7) | 0.911 |
| Embryos transferred | 1.6 ± 0.6 | 1.6 ± 0.5 | 0.0 (−0.2, 0.2) | 0.818 |
| TQ embryos transferred | 0.3 ± 0.6 | 0.4 ± 0.6 | −0.1 (−0.3, 0.2) | 0.503 |
| TQ quality embryo (%) | 12.5 ± 27.6 | 10.2 ± 18.7 | 2.3 (−6.3, 10.8) | 0.584 |
Depicted are mean and standard deviations or number and proportions.
Adjusted for center (fixed effects, M–H, z-test).
2PN, pronuclear stage oocytes; d2, day two of embryo development; d3, day three of embryo development; d5, day five of embryo development; TQ, top quality; NA, not applicable.
Survival rates and laboratory outcomes (PP population).
| Gavi®n = 57 | Cryotop®n = 61 | DifferenceGavi®/Cryotop®(95% CI) | P-value | |
|---|---|---|---|---|
| 2PN vitrified (elective) | 9.4 ± 6.1 | 8.8 ± 5.0 | 0.6 (−1.8, 3.0) | 0.771 |
| 2PN vitrified (surplus) | 4.6 ± 2.5 | 4.6 ± 3.2 | 0.0 (−1.4, 1.5) | 0.675 |
| 2PN warmed (total) | 4.7 ± 2.3 | 4.4 ± 1.8 | 0.3 (−0.5, 1.1) | 0.428 |
| Survival rate (%) | 94.0 ± 13.5 | 96.7 ± 9.7 | −1.6 (−4.7, 1.4) | 0.170 |
| 100% survival (n, %)a | 44/57 (77.2) | 53/61 (86.9) | −8.5 (−22.1, 5.1) | 0.220 |
| 0% survival (n, %) | 0/57 (0.0) | 0/61 (0.0) | 0 | NA |
| Embryo transfer (n, %) | ||||
| No transfer | 3/57 (5.3) | 1/61 (1.6) | 3.7 (−3.0, 10.2) | 0.277 |
| d2/d3 | 20/57 (35.1) | 23/61 (37.7) | −2.6 (−20.0, 14.7) | 0.768 |
| d5 | 34/57 (59.6) | 37/61 (60.7) | −1.1 (−18.7, 16.7) | 0.911 |
| Embryos transferred | 1.6 ± 0.6 | 1.6 ± 0.5 | 0.0 (−0.2, 0.2) | 0.818 |
| TQ embryos transferred | 0.3 ± 0.6 | 0.4 ± 0.6 | −0.1 (−0.3, 0.2) | 0.503 |
| TQ quality embryo (%) | 12.5 ± 27.6 | 10.2 ± 18.7 | 2.3 (−6.3, 10.8) | 0.584 |
| Gavi®n = 57 | Cryotop®n = 61 | DifferenceGavi®/Cryotop®(95% CI) | P-value | |
|---|---|---|---|---|
| 2PN vitrified (elective) | 9.4 ± 6.1 | 8.8 ± 5.0 | 0.6 (−1.8, 3.0) | 0.771 |
| 2PN vitrified (surplus) | 4.6 ± 2.5 | 4.6 ± 3.2 | 0.0 (−1.4, 1.5) | 0.675 |
| 2PN warmed (total) | 4.7 ± 2.3 | 4.4 ± 1.8 | 0.3 (−0.5, 1.1) | 0.428 |
| Survival rate (%) | 94.0 ± 13.5 | 96.7 ± 9.7 | −1.6 (−4.7, 1.4) | 0.170 |
| 100% survival (n, %)a | 44/57 (77.2) | 53/61 (86.9) | −8.5 (−22.1, 5.1) | 0.220 |
| 0% survival (n, %) | 0/57 (0.0) | 0/61 (0.0) | 0 | NA |
| Embryo transfer (n, %) | ||||
| No transfer | 3/57 (5.3) | 1/61 (1.6) | 3.7 (−3.0, 10.2) | 0.277 |
| d2/d3 | 20/57 (35.1) | 23/61 (37.7) | −2.6 (−20.0, 14.7) | 0.768 |
| d5 | 34/57 (59.6) | 37/61 (60.7) | −1.1 (−18.7, 16.7) | 0.911 |
| Embryos transferred | 1.6 ± 0.6 | 1.6 ± 0.5 | 0.0 (−0.2, 0.2) | 0.818 |
| TQ embryos transferred | 0.3 ± 0.6 | 0.4 ± 0.6 | −0.1 (−0.3, 0.2) | 0.503 |
| TQ quality embryo (%) | 12.5 ± 27.6 | 10.2 ± 18.7 | 2.3 (−6.3, 10.8) | 0.584 |
Depicted are mean and standard deviations or number and proportions.
Adjusted for center (fixed effects, M–H, z-test).
2PN, pronuclear stage oocytes; d2, day two of embryo development; d3, day three of embryo development; d5, day five of embryo development; TQ, top quality; NA, not applicable.
Laboratory and pregnancy outcomes
Approximately 60% of patients had blastocyst culture in the two groups. Absolute and relative numbers of top-quality embryos per warmed 2PN oocyte were similar for cleavage stage and blastocyst stage embryos (Table II). Pregnancy and live birth rates were also similar after transfer of a mean number of 1.6 embryos in both groups (Table III).
Embryo transfer protocol and clinical outcomes (PP population).
| Gavi®Elective and Surplusn = 57 | Cryotop®Elective and Surplusn = 61 | Difference % Gavi®/Cryotop®(95% CI)a | Relative Risk Gavi®/Cryotop®(95% CI)a | P-valueb | |
|---|---|---|---|---|---|
| FET: artificial cycle (n/N, %) | 54/57 (94.7) | 58/61 (95.1) | NA | NA | 0.69 |
| FET: natural or modified natural cycles (n/N, %) | 3/57 (5.3) | 3/61 (4.9) | NA | NA | 0.69 |
| Positive serum hCG (n/N, %) | 18/57 (31.6) | 25/61 (41.0) | −8.9 (−26.0, 8.3) | 0.78 (0.48, 1.28) | 0.19 |
Clinical pregnancy (n/N, %) | 11/57 (19.3) | 14/61 (23.0) | −3.8 (−18.2, 10.6) | 0.83 (0.40, 1.73) | 0.40 |
Ongoing pregnancy (n/N, %) | 7/57 (12.2) | 8/61 (13.1) | −0.9 (−13.2, 11.4) | 0.92 (0.37, 2.30) | 0.56 |
Live birth (n/N, %) | 7/57 (12.2) | 8/61 (13.1) | −0.9 (−13.2, 11.4) | 0.92 (0.37, 2.30) | 0.56 |
| Gavi®Elective and Surplusn = 57 | Cryotop®Elective and Surplusn = 61 | Difference % Gavi®/Cryotop®(95% CI)a | Relative Risk Gavi®/Cryotop®(95% CI)a | P-valueb | |
|---|---|---|---|---|---|
| FET: artificial cycle (n/N, %) | 54/57 (94.7) | 58/61 (95.1) | NA | NA | 0.69 |
| FET: natural or modified natural cycles (n/N, %) | 3/57 (5.3) | 3/61 (4.9) | NA | NA | 0.69 |
| Positive serum hCG (n/N, %) | 18/57 (31.6) | 25/61 (41.0) | −8.9 (−26.0, 8.3) | 0.78 (0.48, 1.28) | 0.19 |
Clinical pregnancy (n/N, %) | 11/57 (19.3) | 14/61 (23.0) | −3.8 (−18.2, 10.6) | 0.83 (0.40, 1.73) | 0.40 |
Ongoing pregnancy (n/N, %) | 7/57 (12.2) | 8/61 (13.1) | −0.9 (−13.2, 11.4) | 0.92 (0.37, 2.30) | 0.56 |
Live birth (n/N, %) | 7/57 (12.2) | 8/61 (13.1) | −0.9 (−13.2, 11.4) | 0.92 (0.37, 2.30) | 0.56 |
Adjusted for center (Mantel–Haenszel, fixed-effects model).
One-sided Fisher’s exact test.
FET, frozen-thawed embryo transfer; NA, not applicable; hCG, human Chorionic Gonadotropin.
Embryo transfer protocol and clinical outcomes (PP population).
| Gavi®Elective and Surplusn = 57 | Cryotop®Elective and Surplusn = 61 | Difference % Gavi®/Cryotop®(95% CI)a | Relative Risk Gavi®/Cryotop®(95% CI)a | P-valueb | |
|---|---|---|---|---|---|
| FET: artificial cycle (n/N, %) | 54/57 (94.7) | 58/61 (95.1) | NA | NA | 0.69 |
| FET: natural or modified natural cycles (n/N, %) | 3/57 (5.3) | 3/61 (4.9) | NA | NA | 0.69 |
| Positive serum hCG (n/N, %) | 18/57 (31.6) | 25/61 (41.0) | −8.9 (−26.0, 8.3) | 0.78 (0.48, 1.28) | 0.19 |
Clinical pregnancy (n/N, %) | 11/57 (19.3) | 14/61 (23.0) | −3.8 (−18.2, 10.6) | 0.83 (0.40, 1.73) | 0.40 |
Ongoing pregnancy (n/N, %) | 7/57 (12.2) | 8/61 (13.1) | −0.9 (−13.2, 11.4) | 0.92 (0.37, 2.30) | 0.56 |
Live birth (n/N, %) | 7/57 (12.2) | 8/61 (13.1) | −0.9 (−13.2, 11.4) | 0.92 (0.37, 2.30) | 0.56 |
| Gavi®Elective and Surplusn = 57 | Cryotop®Elective and Surplusn = 61 | Difference % Gavi®/Cryotop®(95% CI)a | Relative Risk Gavi®/Cryotop®(95% CI)a | P-valueb | |
|---|---|---|---|---|---|
| FET: artificial cycle (n/N, %) | 54/57 (94.7) | 58/61 (95.1) | NA | NA | 0.69 |
| FET: natural or modified natural cycles (n/N, %) | 3/57 (5.3) | 3/61 (4.9) | NA | NA | 0.69 |
| Positive serum hCG (n/N, %) | 18/57 (31.6) | 25/61 (41.0) | −8.9 (−26.0, 8.3) | 0.78 (0.48, 1.28) | 0.19 |
Clinical pregnancy (n/N, %) | 11/57 (19.3) | 14/61 (23.0) | −3.8 (−18.2, 10.6) | 0.83 (0.40, 1.73) | 0.40 |
Ongoing pregnancy (n/N, %) | 7/57 (12.2) | 8/61 (13.1) | −0.9 (−13.2, 11.4) | 0.92 (0.37, 2.30) | 0.56 |
Live birth (n/N, %) | 7/57 (12.2) | 8/61 (13.1) | −0.9 (−13.2, 11.4) | 0.92 (0.37, 2.30) | 0.56 |
Adjusted for center (Mantel–Haenszel, fixed-effects model).
One-sided Fisher’s exact test.
FET, frozen-thawed embryo transfer; NA, not applicable; hCG, human Chorionic Gonadotropin.
Procedure duration
The mean total procedure duration (vitrification and warming) required for vitrifying and warming of 2PN oocytes was higher in the Gavi® group than in the Cryotop® group (81 ± 39 min vs 47 ± 15 min, mean difference 34 min, 95% CI 19 to 48) (Table IV). Most of the procedure duration increase originates from the warming step which took approximately double the time as compared to the manual procedure.
Procedure characteristics.
| Gavi® | Cryotop® | Difference (95% CI) | P-value | |
|---|---|---|---|---|
| Vitrification duration (min) | 37.0 ± 23.8 | 26.5 ± 10.8 | 10.5 (3.5, 17.5) | 0.037 |
| Warming duration (min) | 37.4 ± 21.0 | 18.9 ± 12.1 | 18.5 (11.06, 25.8) | <0.001 |
| Total durationa (min) | 80.6 ± 38.7 | 46.9 ± 15.0 | 33.7 (19.4, 48.1) | <0.001 |
| Vitrification duration per 2PN vitrified (min) | 6.6 ± 5.2 | 5.2 ± 3.4 | 1.4 (−0.1, 3.0) | 0.039 |
| Warming duration per 2PN warmed (min) | 9.2 ± 7.1 | 4.8 ± 3.0 | 4.4 (2.1, 6.7) | <0.001 |
| Gavi® | Cryotop® | Difference (95% CI) | P-value | |
|---|---|---|---|---|
| Vitrification duration (min) | 37.0 ± 23.8 | 26.5 ± 10.8 | 10.5 (3.5, 17.5) | 0.037 |
| Warming duration (min) | 37.4 ± 21.0 | 18.9 ± 12.1 | 18.5 (11.06, 25.8) | <0.001 |
| Total durationa (min) | 80.6 ± 38.7 | 46.9 ± 15.0 | 33.7 (19.4, 48.1) | <0.001 |
| Vitrification duration per 2PN vitrified (min) | 6.6 ± 5.2 | 5.2 ± 3.4 | 1.4 (−0.1, 3.0) | 0.039 |
| Warming duration per 2PN warmed (min) | 9.2 ± 7.1 | 4.8 ± 3.0 | 4.4 (2.1, 6.7) | <0.001 |
Depicted are mean and standard deviations and mean difference with 95% confidence intervals of procedure durations in minutes.
Total duration was calculated only for those patients with vitrification and warming, who had on average eight 2PN oocytes vitrified.
min, minutes; 2PN, pronuclear stage oocytes.
Procedure characteristics.
| Gavi® | Cryotop® | Difference (95% CI) | P-value | |
|---|---|---|---|---|
| Vitrification duration (min) | 37.0 ± 23.8 | 26.5 ± 10.8 | 10.5 (3.5, 17.5) | 0.037 |
| Warming duration (min) | 37.4 ± 21.0 | 18.9 ± 12.1 | 18.5 (11.06, 25.8) | <0.001 |
| Total durationa (min) | 80.6 ± 38.7 | 46.9 ± 15.0 | 33.7 (19.4, 48.1) | <0.001 |
| Vitrification duration per 2PN vitrified (min) | 6.6 ± 5.2 | 5.2 ± 3.4 | 1.4 (−0.1, 3.0) | 0.039 |
| Warming duration per 2PN warmed (min) | 9.2 ± 7.1 | 4.8 ± 3.0 | 4.4 (2.1, 6.7) | <0.001 |
| Gavi® | Cryotop® | Difference (95% CI) | P-value | |
|---|---|---|---|---|
| Vitrification duration (min) | 37.0 ± 23.8 | 26.5 ± 10.8 | 10.5 (3.5, 17.5) | 0.037 |
| Warming duration (min) | 37.4 ± 21.0 | 18.9 ± 12.1 | 18.5 (11.06, 25.8) | <0.001 |
| Total durationa (min) | 80.6 ± 38.7 | 46.9 ± 15.0 | 33.7 (19.4, 48.1) | <0.001 |
| Vitrification duration per 2PN vitrified (min) | 6.6 ± 5.2 | 5.2 ± 3.4 | 1.4 (−0.1, 3.0) | 0.039 |
| Warming duration per 2PN warmed (min) | 9.2 ± 7.1 | 4.8 ± 3.0 | 4.4 (2.1, 6.7) | <0.001 |
Depicted are mean and standard deviations and mean difference with 95% confidence intervals of procedure durations in minutes.
Total duration was calculated only for those patients with vitrification and warming, who had on average eight 2PN oocytes vitrified.
min, minutes; 2PN, pronuclear stage oocytes.
Staff convenience
Staff convenience as assessed in eight operators by a questionnaire was lower for the Gavi® system (Supplementary Fig. S2). The majority of respondents (five out of eight) preferred the Cryotop® method. The majority of respondents found that the Gavi® method was not easier to use (six out of seven) and not time saving (six out of seven). Four out of eight respondents would not recommend the Gavi® method to a colleague and four out of eight respondents were not sure whether or not to recommend it.
Discussion
Most of the processes in the IVF laboratory are performed manually and they are hence dependent on the level of hands-on experience of the operating lab staff. Some impact of inter- and intra-individual variation of the performance and outcome is, however, still to be expected even in a well-controlled setting. A number of efforts have been made to promote standardization, consistency and efficiency of different and crucial steps of ART procedures by the development and/or the introduction of new (automated) technologies such as non-subjective sperm analysis and selection (Ainsworth et al., 2005; Nosrati et al., 2014), time-lapse based monitoring of embryo development and selection (Kirkegaard et al., 2012), non-invasive metabolic evaluation of embryo viability (Vergouw et al., 2008), media replacement and oocyte denudation by microfluidics (Zeringue and Beebe, 2004; Swain et al., 2013) and oocyte positioning (Keefe et al., 2003; Saadat et al., 2018), to name a few. However, the implementation of automation into established laboratory routine should be preceded by outcome and impact analyses in the context of a clinical trial. The Gavi® device, the first semi-automated vitrification system, was introduced to the European market in 2015, yet no data for procedure characteristics and outcomes investigated in a controlled trial setting had been available.
The presented data of this study demonstrate that the performance of the semi-automated closed Gavi® method and the manual open Cryotop® method is comparable with regard to the primary outcome measure cryosurvival of PN stage oocytes. To note, the observed zygote survival rates within this study, 94.0% (±13.5) and 96.7% (±9.7) in the Gavi® and the Cryotop® group, are higher than the benchmark value of 85% for zygote survival suggested by Alpha Scientists (Alpha Scientists in Reproductive Medicine, 2012). Similarly, no statistically significant or clinically relevant differences in laboratory or clinical outcomes were observed in descriptive statistics on secondary endpoints. It is noteworthy, however, that the absence of blinding in this trial and the freedom of protocol choice when transferring embryos after warming have potential negative implications on the robustness of the findings of no outcome differences for pregnancy and live birth achievement. Future trials should therefore consider using a blinded design, further and more stringent stipulations on clinical protocols and will also need to be performed on much larger samples for testing differences in live birth outcome, child health, etc.
The efficacy of the Gavi® device demonstrated herein is in line with preliminary pre-clinical data and with the available evidence from non-randomised studies. Roy et al. assessed the efficacy of the Gavi® system for mouse zygotes and embryos as well as for human blastocysts, and it was demonstrated that the semi-automated vitrification method has the potential to produce comparable results in terms of recovery rate for mouse zygotes and equivalent results for mouse embryos and mouse blastocysts as compared to the Cryotop® system. Despite a significantly lower survival rate of mouse zygotes being achieved after automated processing with the Gavi® system, blastocyst formation was comparable between both vitrification methods. Furthermore, comparable development of vitrified-warmed human blastocysts up to the fully hatched blastocyst stage on Day 6 was observed (Roy et al., 2014). According to a pilot study in the human, comparing the clinical outcome after vitrification and warming using both systems, the first pregnancies from Gavi® vitrified-warmed biopsied human blastocysts were achieved (Hobson et al., 2016). A recently published retrospective study comparing the two vitrification approaches reported equivalent survival rates for human blastocysts of 98.6% after semi-automated vitrification and 99.3% when manually vitrified. Similarly, clinical outcomes in terms of pregnancy rate (33.4% vs 34.6%) and miscarriage rate (24.8% vs 22.2%) after transfer of vitrified-warmed assisted-hatching laser-treated blastocysts were reported (Miwa et al., 2020).
Although closed vitrification devices are associated with lower cooling and warming rates, it has been suggested that similar cryosurvival rates for embryos and blastocysts can be achieved when compared to open vitrification carrier systems (Kuwayama et al., 2005; Chen et al., 2013; Desai et al., 2013; Panagiotidis et al., 2013). The present study corroborates this notion on cryosurvival of 2PN oocytes in the context of a semi-automated device with the evidence generated being supportive of only negligible outcome differences (exemplified by the −4.7% lower bound CI of the difference in mean 2PN oocyte survival rate). It is likely that our findings hold true for cleavage stage embryos and blastocysts, since cryosurvival is generally similar between these entities. A non-inferiority study design was chosen for this trial, since survival rates are reported to be >95% in open vitrification systems (Golakov et al., 2018) and the claimed benefits of the Gavi® system (short learning curve, low human error rate, less inter-operator variations and low contamination risk) can theoretically come at the expense of decreased efficacy in survival rate though within tight margins. Of note, the Gavi® system does not allow flexibility in adapting the vitrification protocol. Successful vitrification of oocytes, zygotes and embryos at various developmental stages depends on optimal dehydration and penetration of cryoprotectants. Whereas the Cryotop® method allows the observation of shrinkage and re-expansion of cells occurring during exposure to cryoprotectants under a microscopic view and if required adaption of the incubation time, the Gavi® does not provide the possibility of visual control, which may be considered a relevant flaw of the semi-automated vitrification system.
To also include patients with no confirmed post-warming survival of 2PN oocytes or potential total loss of 2PN oocytes, a per-protocol analysis was used for the primary outcome as well for the secondary clinical outcomes. A per-protocol analysis is also the recommended approach for non-inferiority trials to arrive at a conservative inference on the primary hypothesis (Dunn et al., 2018). Of note, a randomised trial on the primary outcome survival rate post warming will inherently show high attrition between subjects at randomization and subjects in whom the primary outcome can be assessed. However, since this attrition is neither systematic nor directly linked to the primary outcome, a biased estimate of the measured effect is unlikely. It is also noteworthy that recruitment into this trial was stopped prematurely by the principal investigator who had information about the randomised patient numbers in the three centers at the two strata and who decided to stop recruitment given the high proportion of freeze-all cycles already recruited into the trial. The benefit of doing so lies in saving time (and resources) until complete assessment of all outcomes given that the maximum theoretical time span from randomization of the last subject to end of study was still approximately 66 weeks. Stopping the trial prematurely, however, could have resulted in a biased study conclusion, e.g. a premature stopping of the trial at a time when the data looked in favor of the hypothesis to be tested. This, however, is unlikely for at least three reasons: first, at the time of stopping further recruitment and randomization, neither the total number of patients with warming nor the complete outcome could have been available to the investigators; second, the number of patients necessary to test the hypothesis had a priori been calculated and was met (the number planned to randomise was based on assumptions necessary to reach that number); third, given the flow of patients through the trial, a continuation of recruitment would have resulted in only about 10–12 additional observations per group, a number unlikely to be able to invalidate the finding of non-inferiority given the highly similar outcome documented at the end of study.
A noticeable aspect of the trial is the rather low number of top-quality embryos available for transfer. Embryo scoring was performed according to the Alpha/ESHRE consensus on embryo assessment. Considering that only Score 1 embryos were classified top-quality, whereas in other studies embryos of top or good quality were often defined by other criteria (Guerif et al., 2009; Miwa et al., 2020), our strict approach might be a possible explanation for the low incidence of top-quality embryos. Nevertheless, the live birth rate per-protocol of 12–13% was also rather low and compares unfavorably with the German IVF registry vitrified-warmed ET cycle live birth rate of 18–19% per started vitrified-warmed embryo-transfer cycle (Blumenauer et al., 2020). One potential explanation could be the very conservative number of 2PNs warmed per attempted vitrified-warmed ET observed herein, which may relate to the context of a controlled trial and the tendency to stick to a rather strict interpretation of the German Embryo Protection Act (Ziller, 2017) by the involved investigators. Although a 30–40% positive hCG rate in the per-protocol population was observed in this trial, approximately 60–70% of positive hCG tests did not result in a live born infant, perhaps as a result of the transfer of embryos of poor developmental potential. Since these findings, however, are symmetrical between groups, a causal link to the type of vitrification procedure is highly unlikely.
Beyond the primary outcome cryosurvival of 2PN oocytes, we aimed at measuring adverse events, such as device failure or total loss of vitrified-warmed PN stage oocytes at warming as well as procedure duration and staff convenience. During the study period, a problem with the automated sealing unit occurred. While performing a mandatory maintenance check, carried out by operating staff, it was noticed that the Gavi® pods were not sealed properly. The technical problem was promptly fixed and had no influence on the course or outcome of the study. Two patients randomised to the Gavi® procedure could not be treated as intended, because not enough medium cartridges were available on that day, which was an organizational issue.
The staff convenience with the Gavi® system was remarkably low and this was consistent across centers. Deficiencies in evaluating the staff convenience, however, need to be addressed, as we could not use a validated tool to measure different aspects of product and procedure satisfaction, respectively. It is also not necessarily uncommon that a new technology is met with skepticism by operators who are used to their routine procedures. Our questionnaire results are therefore to be taken with caution.
However, a robust finding of our trial is the increased time required for both the Gavi® vitrification and warming procedure, with a combined mean duration increase of 34 min and a mean duration increase of 4.4 min per unit when compared to the Cryotop® vitrification/warming method. Based on our data, the claimed benefit of saving hands-on time with the Gavi® (Roy et al., 2014) cannot be confirmed. Although up to eight unfertilized oocytes or 2PN oocytes can be vitrified simultaneously with the Gavi®, a funneling effect may come into play when a large number of entities must be vitrified in case of several freeze all or fertility preservation cycles accumulating on a single day. However, an increased overall procedure duration time may be less relevant in settings were fewer entities such as blastocysts and/or embryos are vitrified and in settings with a low proportion of vitrified-warmed ET cycles overall. Yet, since personnel cost is one of the main contributors to total IVF service provision costs, an increased overall time required for vitrification and warming procedures may have relevant cost implications, the extent of which can be modelled in different settings based on the present trial findings.
Conclusion
Survival of 2PN oocytes post-warming, as well as embryo development and clinical outcomes are similar between the semi-automated closed the Gavi® vitrification method and the manual open Cryotop® vitrification method. With increasing requirements for standardization of procedures and tissue safety, the Gavi® automat may constitute a suitable alternative technology to the established manual open vitrification technique given the equivalent clinical outcomes demonstrated herein.
Data availability
The data underlying this article will be shared on reasonable request to the corresponding author.
Acknowledgments
Our thanks go to Nils Rutsch, Jens Popken and Heidrun Muschalla for their support in the Gavi® vitrification/warming set-up and staff training in the three study centers, as well as to Prof. Inke Koenig of the Institute for Medical Biometry and Statistics at the University of Luebeck for statistical advice.
Authors’ roles
J.H. was responsible for the study design, study initiation, trial audit, data retrieval, data analysis and interpretation, manuscript drafting and finalizing. G.G. was responsible for the study design, study initiation and supervision, data analysis and interpretation, manuscript drafting and finalizing. R.B. was responsible for the trial audit and data retrieval and analysis, and participated in manuscript drafting and finalizing. N.S.M., S.S., B.S., M.D., A.S.M., K.N., F.G. and S.O. were responsible for data acquisition throughout the trial. All authors critically reviewed the manuscript and approved the final version.
Funding
The trial received no direct financial funding. The Gavi® instrument, Gavi® consumables and staff training was provided for free by the distributor (Merck, Darmstadt, Germany) during the study period. The manufacturer of the Gavi® instrument had no influence on study protocol, study conduct, data analysis, data interpretation or manuscript writing.
Conflict of interest
J.H. has received honoraria and/or non-financial support by Ferring, Merck and Origio. G.G. has received honoraria and/or non-financial support by Abbott, Ferring, Finox, Gedeon Richter, Guerbet, Merck, MSD, ObsEva, PregLem, ReprodWissen GmbH and Theramex. The remaining authors have no conflict of interest to declare.
