Day 2 versus day 3 embryo transfer in poor responders: a prospective randomized trial
Article Outline
Day 2 embryo transfer has been suggested as a method to improve pregnancy rates in poor responders compared with day 3 transfer. Our prospective randomized controlled trial does not show a difference in outcomes based on day of embryo transfer.
Key Words: IVF, poor responders, embryo transfer
The incidence of poor response to controlled ovarian stimulation is reported to be from 9% to 24% 1, 2, 3. The optimal management of poor responders remains a challenge despite a multitude of studies dedicated to improving outcomes in this patient population (4). Two studies have shown a possible benefit of an embryo transfer on day 2 compared with day 3 in poor responders, but they conclude that their findings are preliminary 5, 6. The first, a retrospective cohort study, found a higher chance of ongoing pregnancy (46.2% vs. 28.6%; P=.05) in poor responders aged <40 years with an embryo transfer on day 2 instead of day 3. The second, a prospective trial of poor responders randomized to day 2 versus day 3 embryo transfer, showed a higher clinical pregnancy rate (CPR) in the day 2 group (37.2% vs. 21.4%; P<.05). We wanted to investigate the efficacy of this treatment at our center before considering implementation of day 2 transfer as the standard of care for our poor responders. We conducted a prospective randomized trial to evaluate whether the day of embryo transfer (day 2 or day 3) affects clinical pregnancy rates in poor responder patients.
This trial was conducted in a single university IVF program from January 1, 2007, to March 15, 2009, and Institutional Review Board approval was obtained. All patients undergoing fresh autologous IVF treatment were considered for entry into the trial. Patients were approached and informed about the trial before or during their IVF cycle and gave informed consent before or at the time of oocyte retrieval. Patients were randomized by a computer-generated sequence program to either day 2 or day 3 embryo transfer at the time of fertilization, and allocation to treatment group was concealed in an opaque envelope. Neither patients nor providers were blinded to the intervention.
Inclusion criteria for randomization were at least one fertilized oocyte and an intent to transfer all available embryos. Patients planning preimplantation genetic diagnosis were not consented and those with no oocytes retrieved or no fertilization may have been consented but were not randomized. Patients were not compensated for their participation. Each patient was allowed entry into the trial once.
Patients received one of four standard ovarian stimulation protocols: microdose lupron (flare), antagonist, luteal down-regulation (long), or estrogen priming with antagonist. Decision on protocol was left to the preferences of the patients' physician, and choice was based on patient characteristics and history. Details of these protocols have been described previously 7, 8. A dose of 10,000 U hCG was administered when at least one follicle reached an average diameter of ≥17 mm. Oocyte retrieval was performed by transvaginal ultrasound guidance 35 hours after hCG administration.
Fertilization of the oocytes was achieved with standard insemination or intracytoplasmic sperm injection (ICSI). The oocytes were examined for fertilization status 16–18 hours after IVF or ICSI. The zygotes with two pronuclei were cultured for 24–48 hours in Sage cleavage medium (Cooper Surgical, Trumbull, CT) with 10% serum protein substitute (SPS; Irvine Scientific, Santa Ana, CA).
On the day of embryo transfer, embryos for assisted hatching (AH) were placed in phosphate-buffered saline solution (PBS) with 10% SPS. AH was accomplished by using the Zilos-tk laser (Hamilton Thorne Biosciences, Beverly, MA) to make a hole in an area of the zona pellucida between blastomeres.
Transabdominal ultrasound–guided embryo transfer on day 2 or day 3 after oocyte retrieval was performed using a Tefcat or Echotip Softpass catheter (Cook Ob/Gyn, Spencer, IN). All patients received progesterone supplementation with vaginal suppositories (200 mg three times a day) starting on the evening after oocyte retrieval.
A pregnancy test was considered to be positive if the β-hCG level was >5 mIU/mL 10–11 days after embryo transfer. Clinical intrauterine pregnancy (IUP) was defined as gestational sac seen on transvaginal sonogram at 6–7 weeks' gestational age. Implantation rate was defined as number of gestational sacs seen on transvaginal sonogram per number of embryos transfered. Spontaneous pregnancy loss (SAB) was defined as loss of pregnancy after clinical IUP. Biochemical pregnancy was defined as an initially positive β-hCG that declined with serial measurements or no gestational sac on transvaginal sonogram. Ongoing pregnancy was defined as pregnancy beyond 12 weeks' gestation. Live birth data were collected by contacting patients. The primary outcome measured was CPR, and secondary outcomes measured were implantation rate, biochemical pregnancy rate, SAB rate, and live birth/ongoing pregnancy rates.
The two groups were compared with univariate analysis using Student t test for continuous variables and chi-square test for dichotomous variables as appropriate. A P value of <.05 was considered to be significant. A power calculation was done before patient recruitment began. We needed at least 122 patients in each group to show an increase in clinical pregnancy from 15% to 30% with 80% power and 0.05 alpha error. Our poor responder patient population had a 15% clinical pregnancy rate with day 3 embryo transfer (9), and we felt that an increase in clinical pregnancy rate to 30% with day 2 embryo transfer would be enough of an improvement to change our center's standard day of transfer in this population. The earlier prospective trial by Bahceci et al. (6) also used an improved pregnancy rate of 15% to 30% in their power calculation.
A total of 386 patients were identified as possible poor responders (by age, high cycle day 3 (CD3) FSH level, or history of poor response to ovarian stimulation) and considered for entry into the trial. Seventy-nine patients declined entry, and 307 patients were consented before or at the time of oocyte retrieval. Fifty-six patients were not randomized: 11 patients declined randomization after being consented, and 18 patients canceled the oocyte retrieval and converted their IVF cycle to an intrauterine insemination. An additional 27 patients had no embryos on the day of fertilization check: 8 had no oocyte retrieved, and 19 had no fertilization. Of the 251 patients who had at least one fertilized oocyte and were randomized, 3 patients had no cleavage and therefore had no embryo transfer. These patients were included in the analysis as intention to treat (1 patient in the day 2 group and 2 patients in the day 3 group).
The two groups were similar in patient characteristics. In addition to female age and CD3 FSH level (Table 1), the two groups (day 2 vs. day 3, respectively) were similar in body mass index (23.6 vs. 23.5; P=.9), number of previous IVF cycles (1.5 vs. 1.3; P=.4), obstetric history (gravida 0: 43% vs. 44% [P=.7]; para 0: 74% and 73% [P=.7]; history of one or more miscarriages: 28% and 30% [P=.7]), indication for IVF treatment (diminished ovarian reserve only: 47% vs. 45% [P=.8]; endometriosis: 11% vs. 17% [P=.2]; male factor: 32% vs. 28% [P=.5]; tubal disease: 7% vs. 9% [P=.4]; uterine factor: 3% vs. 4% [P=.8]), male partner age (40.8 and 40.9; P=.9), and ethnicity of the female patients (Caucasian 46% and 47% [P=.8], Asian 39% and 38% [P=.9], Hispanic 4% and 2% [P=.2], and Indian 11% and 13% [P=.6]).
Table 1. Day 2 vs. day 3 embryo transfer in poor responders.
| Day 2n = 123 | Day 3n = 128 | P value | |
|---|---|---|---|
| Female age, y | 39.9 ± 3.0 (32.2–45.3) | 39.2 ± 4.0 (29.5–48.4) | 0.1 |
| Cycle day 3 FSH | 11.0 ± 5.4 | 10.2 ± 5.6 | 0.3 |
| Endometrial thickness, mm | 9.6 ± 2.0 | 9.4 ± 1.6 | 0.5 |
| Oocytes retrieved | 4.0 ± 2.2 | 4.2 ± 1.6 | 0.2 |
| Intracytoplasmic sperm injection | 58% | 49% | 0.2 |
| Embryos Transfered | 2.1 ± 2.8 | 2.4 ± 2.6 | 0.1 |
| Positive hCG | 19.5% | 25% | 0.3 |
| Biochemical pregnancy rate | 3.3% | 8.6% | 0.1 |
| Clinical pregnancy rate | 15.4% | 16.4% | 0.8 |
| Implantation rate | 9.0% | 8.9% | 0.9 |
| Spontaneous pregnancy loss | 26.3% | 19.0% | 0.6 |
| Live birth/ongoing pregnancy rate | 12.3% | 12.7% | 0.8 |
IVF cycle characteristics were also similar between the two groups. In addition to the characteristic listed in the table, the two groups (day 2 vs. day 3, respectively) were similar in stimulation protocol (long 7% vs. 8% [P=.7], antagonist 57% vs. 61% [P=.5], flare 30% vs. 30% [P=.8], and estrogen priming 6% vs. 1% [P=.1]), total amount of gonadotropins used (6,218 IU vs. 6,072 IU; P=.4), number of mature follicles on day of hCG trigger (5.0 vs. 5.3; P=.1), percentage of patients with AH (88% vs. 82%; P=.2), and fertilization rate (60% vs. 63%; P=.6).
IVF outcomes were similar between the two groups and are listed in Table 1. All ongoing pregnancies and live births were singletons. We conducted subgroup analyses in an attempt to identify a patient population that may benefit from either day 2 or day 3 embryo transfer. We did not find a difference in CPR between treatment groups in patients analyzed by the following subgroups: female age ≥40 years versus <40 years, CD3 FSH level ≥12 mIU/ML versus lower CD3 FSH level, patients with only one embryo to transfer, patients with a history of two or more failed IVF cycles, patients with different ovarian stimulation protocols, or patients with or without ICSI or assisted hatching.
Assisted reproduction has impressive success rates in patients with various indications for IVF as long as ovarian stimulation results in a good number of oocytes (10). Patients with poor response to controlled ovarian stimulation have a worse prognosis, and many treatment options for improving success with IVF have been proposed (4). Two previously published studies have shown a benefit of day 2 embryo transfer in poor responders. The first study, by Shen et al., 2006 (5), examined the influence of day of embryo transfer in poor responders, defined as patients who transfered all available embryos. This retrospective study comparing outcomes during two different time periods found that patients <40 years old (mean ages 36.3 and 35.9) had a higher chance of ongoing pregnancy (46.2% vs. 28.6%; P=.05) and a lower chance of SAB (6.3% vs. 31.0%; P<.05) if they had their embryos transferred on day 2 compared with day 3, respectively. The IVF outcomes did not differ in patients >40 years old. The second study, by Bahceci et al., 2006 (6), was a prospective randomized trial comparing IVF outcomes between day 2 or day 3 embryo transfer in poor responders, defined as those with fewer than five follicles measuring >13 mm at the end of ovarian stimulation. The patients with a day 2 embryo transfer (average age 36.5 years) had a significantly higher chance of ongoing pregnancy per oocyte retrieval compared with those with a transfer on day 3 (average age 36.6 years; 27.7% vs. 16.2%, respectively; P=.02).
Unlike the earlier trials, we did not find a benefit for day 2 transfer in poor responder patients. This discrepancy may be due to differences between the study design and patient populations in the studies. The lower overall pregnancy rate in our study compared with the two earlier studies may be due to the fact that our patient population was on average 3 years older. The work by Shen et al. (5) was a retrospective study comparing outcomes during two different time periods, allowing for some bias and confounders which the present prospective randomized trial should limit.
Although the study by Bahecci et al. (6) was also a prospective randomized trial, it differed from ours in three distinct ways. First, we defined poor responders as patients with a limited number of embryos to transfer instead of a limited number of mature follicles before retrieval. There is no clear definition of poor responder in the IVF literature (4), and we think a limited number of embryos is more clinically relevant and applicable to practice compared with follicles. Second, we randomized patients at the time of fertilization report and only included patients with at least one fertilized oocyte instead of before retrieval, because this is more clinically relevant and would allow for a high percentage of patients receiving the intended treatment (99% in our trial and only 92% in the trial by Bahecci et al.). Third, the mean numbers of embryos transfered were similar in the day 2 and day 3 transfer groups (2.1 vs. 2.4; P=.1) in our trial, but they differed in the study by Bahecci et al. (2.0 vs. 1.7; P=.003). The significantly higher number of embryos transfered in the day 2 patients may account for the increased ongoing pregnancy rates in that intervention group compared with day 3 group in the Bahecci et al. trial.
Our data is more consistent with a multitude of studies looking at the effect of day 2 embryo transfer on IVF success in good-prognosis patients. A pooled analysis of multiple randomized controlled trials in the general IVF population found insufficient evidence to suggest an improved live birth rate with embryo transfer on day 2 compared with day 3 (odds ratio 1.07, 95% confidence interval 0.84 to 1.37) (11). In contrast, earlier studies suggest that earlier transfer might benefit poor-prognosis patients because patients with poor embryo quality may be more susceptible to the detriments of the in vitro environment 5, 6. However, the present prospective randomized trial in patients with a limited number of embryos to transfer found no difference in outcome if embryo transfer was on day 2 or day 3. Therefore, we think that day 2 and day 3 embryo transfers should be considered to be similar options in poor responders, depending on clinician and patient preferences.
References
- . Evaluating strategies for improving ovarian response of the poor responder undergoing assisted reproductive techniques. Fertil Steril. 2000;73:667–676
- Comparison of “poor” responders with “good” responders using a standard buserelin/human menopausal gonadotropin regime for in-vitro fertilization. Hum Reprod. 1991;6:918–921
- . Combined gonadotropin releasing homone agonist/human menopausal gonadotropin therapy (GnRH-a/hMG) in normal, high, and poor responders to hMG. J In Vitro Fert Embryo Transf. 1991;8:33–36
- . How to improve the probability of pregnancy in poor responders undergoing in vitro fertilization: a systematic review and meta-analysis. Fertil Steril. 2009;91:749–766
- . Day 2 transfer improves pregnancy outcome in in vitro fertilization cycles with few available embryos. Fertil Steril. 2006;86:44–50
- . Efficiency of changing the embryo transfer time from day 3 to day 2 among women with poor ovarian response: a prospective randomized trial. Fertil Steril. 2006;86:81–85
- Defining human embryo phenotypes by cohort-specific prognostic factors. PLoS ONE. 2008;3:e2562
- . Use of luteal estradiol patch and gonadotropin-releasing hormone antagonist suppression protocol before gonadotropin stimulation for in vitro fertilization in poor responders. Fertil Steril. 2005;84:1023–1026
- Shahine LK, Lathi RB, Baker VB. Oocyte retrieval versus conversion to intrauterine insemination in patients with poor response to gonadotropin therapy. Fertil Steril. Published online April 24, 2009.
- . Combined impact of the number of pre-ovulatory oocytes and cryopreservation on IVF outcome. Hum Reprod. 1991;6:284–289
- . Day three versus day two embryo transfer following in vitro fertilization or intracytoplasmic sperm injection. Cochrane Database Syst Rev. 2004;(2):CD004378
L.K.S. has nothing to disclose. A.A.M. has nothing to disclose. L.M.W. has nothing to disclose. V.L.B. has nothing to disclose. B.B. has nothing to disclose. R.B.L. has nothing to disclose.
PII: S0015-0282(10)01048-4
doi:10.1016/j.fertnstert.2010.06.093
© 2011 Published by Elsevier Inc.
