What is the optimal timing of intracytoplasmic sperm injection (ICSI) after EGG retrieval? A randomized controlled trial

Abstract

Purpose

To determine if oocyte denudation and ICSI at 36.5 versus 39 h post HCG and/or Lupron trigger (2.5 h versus 5 h post-oocyte retrieval) influences fertilization and blastulation rates in good prognosis couples

Methods

We performed a prospective, randomized controlled trial of 12 patients undergoing IVF with ICSI at an academic fertility center, resulting in 206 MII oocytes analyzed. At time of retrieval, patients with more than 10 oocytes retrieved had their oocytes randomized into two groups—one group for oocyte denudation and ICSI at 36.5 h post HCG and/or Lupron trigger and the other group for these procedures at 39 h post HCG and/or Lupron trigger (2.5 and 5 h after oocyte retrieval). Primary outcomes were fertilization and blastulation rates.

Results

No difference was observed in fertilization rate, total blastulation rate, or day of blastulation based on timing of denudation and ICSI (all p > 0.05). Multiple regression analyses for fertilization and blastulation controlling for age and BMI revealed no difference in fertilization based on time of ICSI (p = 0.38, 0.71, respectively). A conditional logistic regression to account for multiple oocytes derived from each patient also found no difference in fertilization or blastulation based on timing of ICSI, even when controlling for age and BMI (p = 0.47, 0.59, respectively).

Conclusion(s)

In good prognosis couples, we observed no difference in fertilization or blastulation rates based on timing of ICSI within the currently accepted 2- to 6-h window post-retrieval based on a 34-h trigger. The oocyte appears to have a physiological tolerance for fertilization during this window of time, and variability in the timing of ICSI during this window is unlikely to have an impact on cycle outcome.

Introduction

Intracytoplasmic sperm injection (ICSI) during in vitro fertilization (IVF) cycles has revolutionized assisted reproduction. ICSI involves the injection of a single sperm into the cytoplasm of oocytes, thus bypassing the zona pellucida and normal fertilization mechanisms, allowing couples with severe male factor infertility to produce genetically related offspring [1–3]. Since the adoption of ICSI into practice in the 1990s, it has been utilized not only for male factor infertility but also for previous fertilization failures with conventional insemination, unexplained infertility, low oocyte yield, advanced maternal age, fertilization after in vitro maturation, fertilization of cryopreserved oocytes, and pre-implantation genetic testing [4]. Indeed, owing to its ability to bypass normal barriers to fertilization, ICSI use in IVF centers continues to rise, with an increase from 36.4% of fresh IVF cycles in 1996 to 76.2% of cycles in 2012 [5]. However, despite nearly 20 years of use, a standardization of the ICSI procedure has not been achieved [3, 6].

Both cytoplasmic maturation and nuclear maturation of the oocyte are important in maximizing the potential for fertilization. Maturation is first induced by the luteinizing hormone (LH) surge in natural cycles or hCG injection in stimulated cycles. Oocyte aspiration is normally scheduled for 34 to 36 h later. Nuclear maturation can be easily assessed in the laboratory by the observation of the first polar body. However, cytoplasmic maturation cannot be assessed visually or biochemically and may not be synchronous with nuclear maturity in stimulated cycles [7, 8]. Moreover, the cumulus and corona cells that surround the oocyte after aspiration are thought to aid in the maturation process and, indeed, some immature oocytes can mature to metaphase II (MII) oocytes in culture. Conversely, a lengthy incubation period of the oocyte-cumulus complex may accelerate the oocyte aging process, diminishing fertilization potential. Thus, the amount of time the oocyte is exposed to its surrounding cellular matrix in culture may affect the oocyte’s potential to produce a viable embryo for transfer. Currently, the consensus for IVF with conventional insemination shows that an incubation period of 2–6 h, prior to insemination, maximizes fertilization and pregnancy rates [9, 10]. However, with insemination, once the sperm are added to the dish containing the oocyte, the timing of sperm binding and penetration is subject to natural mechanisms and not strictly controlled. Conversely, with ICSI, the timing of entry of the sperm into the oocyte is determined artificially. Presently, a strict timeline for the ICSI procedure has yet to be determined or standardized [6].

Prior to injection of sperm, the cumulus cells that surround the oocyte must be removed, both mechanically and enzymatically, a process known as oocyte denudation. The timing of this can be manipulated, and a handful of studies have investigated whether a longer or shorter incubation period of the oocyte with the cumulus-corona cells impacts fertilization rates. Several studies have consistently found that an incubation time of ≥ 2–3 h of the oocyte with its cumulus-corona complex enhances fertilization rates [11–16]. One study found that clinical pregnancy rates were higher when the oocytes incubated with their cumulus-corona complex for ≥ 2–3 h, but no singular best time was found [11]. In contrast, one study did not find any difference in survival, fertilization, or embryo cleavage rates between oocytes denuded and injected 1–2 h after aspiration versus 5–6 h [17]. In terms of timing from denudation to injection, several studies have shown that immediate versus delayed injection after denudation did not appear to influence fertilization rates when performed within the 2–6-h period [18, 19]. Most centers inject sperm immediately after denudation to optimize workflow.

Given the lack of consensus over the timing of important events in the ICSI procedure, namely oocyte denudation, and the widespread adoption of ICSI as a part of assisted reproduction, we sought to investigate if there is a difference in fertilization rates with early (36.5 h post trigger, 2.5 h post oocyte retrieval) and late (39 h post trigger, 5 h post oocyte retrieval) oocyte denudation and immediate sperm injection. We hypothesized that a long in vitro incubation time would improve fertilization and blastulation rates.

Materials and methods

Patient selection

Participants were recruited from the patient population at USC Fertility Center undergoing IVF with ICSI who underwent for the treatment of infertility. Inclusion criteria for female participants included the following: age between 18 and 38 years old at cycle start; serum FSH < 13.5 IU/mL at cycle start; AMH > 1 ng/mL at initial infertility evaluation; ≥ 10 oocytes retrieved. Exclusion criteria for female participants included age over 38 years old; serum FSH > 13.5 at cycle start; AMH < 1 ng/mL; and less than 10 oocytes retrieved. Of note, our center has a BMI requirement of BMI < 43 kg/m² for those pursuing IVF. Thus, all female patients met these criteria for treatment and inclusion in our study. Inclusion criteria for male participants were as follows: age < 50 years old; total motile sperm count > 10 million/mL on baseline semen analysis; no history of varicocele, chemotherapy, or testicular surgery; no history of tobacco use in past 3 months. Exclusion criteria for male participants included age 50 years old; total motile sperm count < 10 million/mL; history of varicocele, chemotherapy, or testicular surgery; history of tobacco use in past 3 months. We have no BMI restrictions for male partners pursuing IVF. Institutional Review Board (IRB) approval was obtained from the University of Southern California (HS-17-00993). Written informed consent was obtained from all participants.

Study design

We performed a prospective, randomized controlled trial with each couple serving as their own control. Prior to study initiation, a statistician created twenty randomization lists with 1:1 ratio of assignments to treatment groups (group 1—2.5 h ICSI; group 2—5 h ICSI) and block size 4. Each list specified random assignment of up to 20 oocytes and was delivered in a separate sealed envelope. For patients with more than 20 oocytes retrieved, additional oocytes above the first 20 were then assigned in 1:1 ratio to the treatment groups, although this only applied to two patients in our study. Only one envelope was used per patient and treatment allocation was concealed until randomization was implemented for a patient. At oocyte retrieval, a numeric code was given to each oocyte in ascending order. The numeric code given to the oocyte was then used to randomize the oocytes to treatment groups based on the randomization card. At 2.5 h after oocyte retrieval, those oocytes assigned to group 1 were denuded by brief exposure (60 s) to hyaluronidase solution (Irvine Scientific, Santa Ana, CA) and repeated pipetting with bore size of 300, 170, 140, and 130 μm (Cook Medical, Vandergrift, PA) to completely remove cumulus and corona cells. ICSI was performed using an Olympus IX73 microscope (Olympus Corporation, Shinjuku City, Japan) with a heated stage (Tokai Hit, Japan) and Narishige micromanipulator (Narishige, Tokyo, Japan) within 30 min of completion of denudation. This same procedure was repeated with oocytes assigned to the 5-h group for denudation and ICSI.

Ovarian stimulation and embryo culture

Controlled ovarian stimulation was performed using injectable gonadotropins (follitropin alfa, EMD Serono; follitropin beta, Merck) or menotropins (Ferring Pharmaceuticals). The endogenous LH surge was suppressed using a gonadotropin-releasing hormone antagonist (ganirelix acetate, Merck; cetrorelix acetate; EMD Serono). When at least two leading follicles reached 18 mm in diameter, oocyte maturation was then accomplished with either BhCG (choriogonadotropin alfa, EMD Serono; chorionic gonadotropin, Ferring), leuprolide acetate, or both. All patients in our study utilized a co-trigger with 2500 IU BhCG and 4 mg of Lupron given in two doses 12 h apart. Oocyte retrieval was then performed 34 h later under transvaginal ultrasound guidance.

Oocyte denudation and ICSI were performed as described above. Embryos were then cultured to the blastocyst stage and vitrified for storage as all patients were planned for a frozen embryo transfer.

Statistical analysis

Our primary outcomes were fertilization and blastulation rates, defined as the number of normally fertilized embryos per oocyte injected and the number of blastocysts per normally fertilized embryo, respectively. For day of blastulation, this was calculated as the number of blastocyst that reached this stage on that specific day divided by the total number of blastocysts from days 5, 6, and 7. Regression analyses conditioned on the patient from whom each oocyte had been retrieved were used to estimate relative rates of these events, treated 2.5 h as the reference value. Multiple and conditional logistic regressions were used to control for covariates (FSH, estradiol, anti-Mullerian hormone, age, and body mass index), to account for multiple oocytes derived from each patient, and to assess for difference in outcomes. A sample size of 200 oocytes was determined by an a priori power calculation to detect rate ratio of 78% or lower in either outcome at a 5% level of nominal significance.

Results

A total of 12 patients were enrolled, resulting in 206 MII oocytes randomized with 105 in group 1 and 101 in group 2. Baseline characteristics of patients enrolled are as described in Table 1. Patients pursued IVF with ICSI for a variety of indications, including polycystic ovarian syndrome (n = 4); unexplained infertility (n = 2 ); same sex couple (n= 4); recurrent pregnancy loss wanting to pursue preimplantation genetic testing for aneuploidy (PGT-A) (n =1); and elective fertility preservation for deferred childbearing (n = 1). All couples had either no success after intra-uterine insemination (IUI) cycles or elected to forgo IUI cycles in lieu of expedited treatment with IVF. All couples had elected to utilize PGT-A. As such, all were to undergo ICSI as part of planned treatment and our center’s standard for those pursuing PGT-A.

Table 1.

Baseline characteristics of patients enrolled

	Median (IQR)
Female age (years)	33.0 (31.0–36.5)
Female BMI (kg/m2)	20.9 (18.9–24.8)
Female day 3 FSH (mIU/mL)	5.8 (4.4–7.5)
Female AMH (ng/mL)	4.8 (3.7–6.5)
Male age (years)	34.5 (31.0–37.3)
Male BMI (kg/m2)	26.3 (23.0–32.0)
Total motile sperm count, pre-wash (million)	28.6 (18.9–45.2)
Total Motile sperm count, post-wash (million)	11.6 (6.2–10.1)

All patients underwent controlled ovarian stimulation using an antagonist protocol. The median number of total oocytes and MII oocytes retrieved per patient was 20.5 (IQR 11.0–19.8) and 17.2 (IQR 11.0–19.8) oocytes, respectively. The median numbers of oocytes per patient in groups 1 and 2 were 8.74 (IQR 6.3–9.8) and 8.42 (IQR 4.25–10.0), respectively.

Fertilization, blastulation, and day of blastulation results are as shown in Fig. 1 and Table 2. There were no differences seen in fertilization rate, blastulation rate, or the day of blastulation based on timing of ICSI. Regression analyses revealed fertilization was less frequent among group 2 oocytes (5 h) with an estimate of relative fertilization rate 0.81 (95% confidence interval [CI]: 0.40–1.63, p=0.55). However, among women for whom any blastulation event occurred, blastulation was more frequent in group 2 oocytes (5 h); estimates of relative blastulation rate were 1.42 (95% CI: 0.81–2.49, p=0.22) overall, and 1.76 (95% CI: 0.84–3.71, p=0.14) among women 35 years of age or younger at the time of retrieval. No relative rate estimate achieved statistical significance. A multiple regression analysis controlling for age, ovarian reserve testing, and BMI found no difference in fertilization or blastulation based on the timing of ICSI (p = 0.38, 0.71, respectively). A conditional logistic regression to account for multiple oocytes derived from each patient also found no difference in fertilization or blastulation based on timing of ICSI, even when controlling for age and BMI (p = 0.47, 0.59, respectively).

Fig. 1.

Fertilization and blastulation rates

Table 2.

Crude rates of fertilizations and blastulation following 2.5 vs 5 h of incubation

	Group 1, 2.5-h incubation (n = 105)	Group 2, 5.0-h incubation (n = 101)	p-value
Fertilization rate (#2PN/#MII)	89/105 (84.8%)	82/101 (81.2%)	0.47
Total blastulation rate (#Blasts/#2PN)	59/89 (66.3%)	62/82 (75.6%)	0.59
Day 5 blastulation (#D5 Blasts/#Blasts)	33/59 (55.9%)	34/62 (54.8%)	0.6
Day 6 blastulation (#D6 Blasts/#Blasts)	22/59 (37.3%)	25/62 (40.3%)	0.74
Day 7 blastulation (#D7 Blasts/#Blasts)	4/59 (6.8%)	3/62 (4.8%)	0.92

Discussion

Since its inception, the utilization of ICSI has transformed male factor infertility and cases of total fertilization failure. The debate over whether ICSI should be performed for other indications, such as low oocyte yield, unexplained infertility, and advanced maternal age, has remained in question. However, in spite of this, rates of ICSI, particularly in cases with non-male factor infertility, continue to increase [5]. As such, determining the best timing for ICSI within the currently accepted 2- to 6-h time frame after oocyte retrieval is paramount to ensure there we are optimizing fertilization and blastulation rates for patients with non-male factor infertility.

To our knowledge, this is the first randomized controlled trial assessing the influence of timing of oocyte denudation and ICSI after egg retrieval on fertilization and blastulation outcomes when performed within the currently accepted 2- to 6-h time window. Moreover, this is one of the few studies using each patient as her own control by splitting her oocytes randomly into two groups. The rate of IVF cycles utilizing ICSI for non-male factor infertility indications will likely increase in the coming years as more patients utilize frozen oocytes, in vitro matured oocytes, or pre-implantation genetic testing. Our work allows embryologists to feel confident in a flexible timing of oocyte denudation and ICSI within the current standards based on laboratory demands and availability without compromising outcomes of good prognosis patients.

Our findings mirror the literature, for which it is common practice to start any fertilization process within a 2- to 6-h time window after egg retrieval. Trounson’s early work showed delayed insemination of oocytes at least 4 h after egg retrieval resulted in a larger proportion of oocytes that fertilized and developed into embryos suitable for transfer [9]. A more recent prospective randomized study by Mizuno et al. examined immediate versus delayed (2 h) denudation of cumulus cells prior to ICSI [20]. Of note, Mizuno et al. utilized a 36–38-h trigger and ICSI was performed 2.5–4 h post oocyte retrieval in this study. Their work found that the percentage of good quality blastocysts available for transfer was higher in those women whose oocytes were allowed to remain in culture with their cumulus cells for 2 h versus those that were denuded immediately after retrieval. They found no difference in fertilization, overall blastulation, pregnancy, and miscarriage rates. Another study by Periera et al. examined time between HCG trigger, oocyte retrieval, cumulus cell removal, and ICSI in patients with a history of complete fertilization failure (CFF) and a subsequent cycle without CFF [21]. They found that compared to cycles with successful fertilization, cycles with CFF had shorter time frames between each of the steps between HCG trigger and oocyte denudation, while maintaining similar overall time frames from trigger to ICSI, suggesting that altering time between steps may be crucial for this subset of patients. Finally, in examining these critical time frames from trigger to ICSI, Garor et al. found that a longer in vitro incubation period could not compensate for a shortened in vivo maturation period, meaning a shorter time from trigger to retrieval [22]. From this, it can be inferred that maintaining an adequate in vivo and in vitro maturation time, meaning overall time from trigger to ICSI, is of utmost important.

Timing from trigger to ICSI is critical as the endogenous LH surge or exogenous HCG trigger initiates a cascade of events that allow for the resumption of meiosis and, ultimately, achievement of metaphase II. Thereafter, sperm penetration will result in extrusion of the second polar body and completion of meiosis. However, if not fertilized, the meiotic spindle will eventually degrade (post-ovulatory aging; [23–25]). Kilani et al. utilized light microscopy to assess spindle formation and, ultimately, degradation at various time points after HCG trigger in oocytes derived from patients less than 38 years old [26]. Their study found that meiotic spindles were seen in 58% of oocytes at 36–36.5 h after HCG trigger with a peak of 96% of oocytes at 39–39.5 h post trigger. Spindles began to degrade after 40.5 h or more after HCG trigger. A similar study by Cohen et al. found that the percentage of oocytes with meiotic spindles was highest at 38 h or more after HCG trigger and that fertilization after ICSI was higher in those with a visible spindle [27]. Our center utilizes a 34-h trigger time frame, meaning ICSI was performed in our study at 36.5 and 39 h post-HCG trigger. Despite the theoretical possibility of a lower percentage of oocytes with meiotic spindle formation for our group 1 oocytes (2.5 h) compared to our group 2 oocytes (6 h), we still found no difference in fertilization and blastulation rates. Thus, at present, it appears that trigger timing of 34 h to retrieval and oocyte denudation and ICSI timing of 2–6 h post retrieval remains acceptable. It is possible that timing of optimal ICSI might be different for eggs from older women; however, this could not be examined within the current study. Altogether, while a pre-incubation with cumulus cells may help with overall oocyte competence for normal embryonic development, there are multiple factors at play. Patient characteristics, ovarian stimulation, and laboratory conditions all ultimately determine the outcome of a cycle.

The main strength of our study is that this is the only randomized controlled trial performed using each patient as her own control at these two time points in the currently accepted window to perform oocyte denudation and ICSI. However, our study has several limitations. Most importantly, our study only included good prognosis male and female patients whose outcomes are likely to be favorable at baseline. It is possible that our findings do not hold true for all infertile or older patients, particularly patients in whom oocyte quality and inherent fertilization potential is diminished. Additionally, centers may differ in time frame from denudation to ICSI. While our center routinely performs oocyte retrieval 34 h after trigger and immediately performs ICSI, other centers may operate differently and our findings may not hold true for these laboratories. To strictly control the timing of sperm entry into each eggs, we had to use ICSI. There is some recent evidence that ICSI fertilization may yield slightly lower fertilization rates than conventional IVF when considering the total number of oocytes collected; however, embryo quality is equivalent between the two methods [28].

Questions remain regarding the utility and efficacy of ICSI in non-male factor infertility cases. From a laboratory workflow perspective, performing oocyte denudation and ICSI at any point within the currently accepted time frame does not appear to compromise fertilization and blastulation outcomes in good prognosis patients. However, the impact on eventual pregnancy outcomes was not assessed in our study. A comprehensive assessment of timing of trigger, oocyte denudation, and ICSI timing on the subsequent impact on implantation and live birth rates is warranted if the current pace of ICSI utilization continues.

In conclusion, our study reaffirms the currently accepted window to perform oocyte denudation and ICSI after oocyte retrieval. The oocyte appears to have a physiological tolerance for fertilization during the currently accepted 2–6-h window of time, and variability in the timing of ICSI during this window is unlikely to have an impact on cycle outcome. As laboratory demands increase, embryologists and clinicians should feel confident in flexible timing for oocyte denudation and ICSI within 2 to 6-h post retrieval.

Declarations

Conflict of interest

The authors declare that they have no conflict of interest.

All procedures performed in this study were in accordance with the ethical standards of the institution (Keck School of Medicine of the University of Southern California Institutional Review Board).