From Oocyte to Blastocyst: A Comparison of IVF Embryo Attrition Rates Based on Ovarian Reserve

INTRODUCTION

In vitro fertilisation (IVF) treatment demands considerable time, emotion, and financial investment from the patient, making it imperative that clinicians set hopeful but realistic expectations prior to treatment start. Effective counselling is particularly important regarding the attrition that occurs with each step of the retrieval and fertilisation process. Attention to the number of oocytes and embryos becomes a crucial objective milestone for patients throughout this journey, and effective provider communication is critical for setting realistic expectations and defining relative success.^[1] Most patient-facing models that explain attrition show average numbers irrespective of fertility diagnosis. While this is helpful for patients with average ovarian reserve, it is often not relevant for patients with robust or diminished ovarian reserve (DOR).

The impact of ovarian reserve on IVF attrition rate remains controversial. Some studies suggest that both high and low ovarian reserve, such as with polycystic ovarian syndrome (PCOS) and DOR, result in inferior oocyte quality that may lead to worse oocyte maturity or blastulation rates.^[2–4] Therefore, it might not be reasonable to expect patients with extremes of ovarian reserve to have the same attrition rates as those with normal ovarian reserve.

There are many variations in oocyte stimulation protocol, often driven by patient medical history, indication for IVF, or previous IVF outcomes. For patients with DOR, clinicians commonly compare high-dose gonadotropin stimulation against stimulation with oral medications and concurrent low-dose gonadotropins (sometimes referred to as minimal stimulation). Existing literature has shown minimal stimulation to be non-inferior to high-dose protocols when considering pregnancy rates.^[5] Advantages of minimal stimulation may include reduced costs, lower medication exposure, decreased risk of hyperstimulation, and increased patient satisfaction.^[6–9] Lower medication exposure with minimal stimulation may better replicate the physiologic environment during folliculogenesis, which has been theorised to benefit oocyte quality.

In this study, we sought to compare fertilisation and blastulation rates for high, normal, and low oocyte stimulation responders. Additionally, we compared attrition rates for patients with DOR who underwent high-dose gonadotropin versus minimal stimulation.

MATERIAL AND METHODS

This retrospective chart review was determined to be IRB-exempt by the Mayo Clinic Institutional Review Board. We included patients aged 18–44 years who underwent IVF treatment for PCOS, unexplained infertility, or DOR between January 1, 2013, and March 23, 2023, at the Mayo Clinic in Rochester, MN. Exclusions included patients under 18 years of age at the time of stimulation start; those who underwent oocyte cryopreservation; those who utilised donor oocytes or gametes; those who had a cleavage stage embryo transfer; those who utilised surgically extracted sperm; those who had a diagnosis of tubal factor, uterine factor, male factor, or genetic causes of infertility; and those who underwent protocols other than antagonist or minimal stimulation.

Patients were stratified into three categories of ovarian reserve: normal ovarian reserve (unexplained infertility), PCOS, and DOR. PCOS was diagnosed based on the 2003 Rotterdam Criteria, with patients meeting at least two of the following: clinical or laboratory evidence of hyperandrogenism, anovulation or oligo-ovulation, and polycystic ovaries on ultrasound.^[10] The diagnosis of DOR was assigned by the treating provider, and all patients in this cohort had an anti-Mullerian Hormone (AMH) level of <1.1 ng/ml, a cutoff for poor responders to IVF previously described by the Bologna criteria.^[11] The DOR cycles were further stratified by antagonist protocol and minimal stimulation protocol.

All patients with PCOS and unexplained infertility underwent the antagonist protocol for ovarian stimulation. A subset of DOR cycles also used the antagonist protocol at the provider’s discretion. In our institute’s antagonist protocol, gonadotropins are started on cycle day three using recombinant Follitropin alpha/beta and menotropins (Menopur^®, Ferring Pharmaceuticals, USA), administered together in the evening. Ganirelix acetate 250 mcg (Fyremadel^®, Ferring Pharmaceuticals, India) is added once daily once E2 reaches or exceeds 300 pg/ml. Ovulation is triggered with either chorionic gonadotropin (Pregynl^®, Organon USA Inc., USA) or leuprolide acetate (Lupron^®, AbbVie Pharmaceuticals, Japan) when at least two follicles meet or exceed 18 mm in size and half of the cohort of follicles measure greater than 15 mm.

The remaining subset of DOR patients underwent the minimal stimulation protocol. In this protocol, stimulation is started on cycle day three with daily generic clomiphene citrate 100 mg. Daily cetrorelix acetate 125 mcg (Cetrotide^®, ASTA Medica, Germany) and every other day menotropins 150 IU (Menopur^®, Ferring Pharmaceuticals, USA) are started when oestradiol levels meet or exceed 300 pg/ml. Chorionic gonadotropin (Pregynl^®, Organon USA Inc., USA) trigger is administered when at least one follicle reaches 18 mm.

RESULTS

Counselling Diagram

Using the data in Table 3, a graphic depiction was created for use in clinical counselling [Figure 1]. For simplicity, the number of eggs retrieved, mature eggs, embryos, and blastocysts was rounded to the nearest whole number. Oocyte maturation, fertilisation, and blastulation rates were rounded to the nearest 5%.

Table 3: Summary statistics based on infertility diagnosis.

	Unexplained (n = 120)	PCOS (n = 143)	DOR, antag stim (n = 45)	DOR, minimal stimulation (n = 77)
Retrieved oocytes (n)	17.7 (11, 21)	24.4 (16, 31)	8.0 (5, 10)	3.0 (2, 4)
Matured oocytes (n)	13.2 (8, 16)	18.8 (12, 24)	5.7 (3, 8)	2.5 (2, 3)
2PN embryos (n)	9.5 (6, 12)	12.5 (8, 16)	3.9 (2, 5)	2.0 (1, 3)
Blastocysts (n)	4.9 (2, 6)	7.4 (4, 9)	2.0 (1, 3)	1.5 (0, 2)
Oocyte maturation rate (%)	77 (68, 87)	78 (68, 92)	72 (52, 88)	85 (75, 100)
Fertilisation rate (%)	73 (62, 87)	68 (59, 82)	69 (53, 86)	74 (50, 100)
Blastulation rate (%)	50 (33, 70)	59 (44, 78)	45 (25, 67)	65 (0, 100)

Data represents mean (25% quantile, median, 75% quantile). PCOS: Polycystic ovarian syndrome, DOR: Diminished ovarian reserve.

Close

Figure 1:

IVF attrition counseling diagram based on ovarian reserve. IVF: In vitro fertilisation, PCOS: Polycystic ovarian syndrome

Export to PPT

DISCUSSION

In our study, no differences in IVF attrition rates were found based on infertility diagnosis for patients who underwent the antagonist protocol. Based on these results, it is reasonable to counsel patients that their blastocyst yield should correlate with oocytes retrieved, regardless of ovarian reserve. Despite similarities in attrition percentages between these groups, a pictorial counselling tool that more accurately portrays both estimates of absolute numbers and percentages will be helpful for personalised patient counselling.

One exception was among DOR patients who underwent minimal stimulation. This group had lower overall attrition compared to the antagonist groups (average numbers: 3.0 oocytes, 2.5 mature, 2.0 fertilised 2PN embryos, and 1.5 blastocysts). Statistically, this was demonstrated by a difference in blastulation rate between DOR-minimal stimulation cycles compared to controls (1.37-fold change, p = 0.04), whereas no difference was noted in blastulation rates between DOR-antagonist cycles and controls (1.0-fold change, p = 0.97). Current literature suggests that minimal stimulation is at least non-inferior to conventional stimulation protocols for DOR patients,^[5] and may have benefits such as reduced risk for ovarian hyperstimulation, thromboembolism, lower medication cost, and improved patient satisfaction.^[6,7] Our study found a higher blastulation rate in minimal stimulation cycles for DOR patients compared to controls, contributing to a growing body of literature favouring this stimulation method for the DOR population. One reason for this may be that lower-dose medications more closely mimic the natural physiology of follicular development, with the potential to create higher-quality oocytes and embryos. Although the absolute mean number of blastocysts was slightly lower in minimal stimulation (1.5) compared to antagonist protocol (2.0) among DOR patients, it is important to consider that minimal stimulation has a significantly lower cost per cycle, and most patients start treatment with the intent of multiple cycles.

The question of whether PCOS patients have lower oocyte quality remains controversial. Several studies show that more oocytes retrieved in PCOS patients do not always translate to more blastocysts or live births compared to non-PCOS patients ^[3,4,12] though studies vary in which step contributes to the greatest attrition. Conversely, other studies have found no difference in fertilisation rates among PCOS patients.^[13] Our study did not find a significant difference in attrition rates among PCOS patients compared to unexplained infertility controls at any step. The variations between studies may be explained by discrepancies in stimulation protocols as well as in the control group among studies. Our study chose a control group of unexplained infertility patients with a normal mean AMH (3.5) compared to PCOS (8.6) to better elicit differences in oocyte potential caused by high response to stimulation. Additionally, our study contained more PCOS patients compared to many prior studies. Of note, in our study, PCOS was diagnosed based on the 2003 Rotterdam Criteria, which has since been updated with the 2023 International Evidence-based Guideline.^[14] In the new guidelines, patients with both irregular cycles and clinical hyperandrogenism meet the diagnosis of PCOS regardless of imaging findings. Furthermore, the updated guidelines state that AMH may be used instead of ultrasound in equivocal cases, but neither imaging nor AMH should be applied to adolescent patients due to poor specificity. In our study, differences in PCOS diagnosis guidelines are unlikely to result in a significant difference since adolescents, who may experience the greatest difference in PCOS diagnosis based on the new guidelines, are excluded as a patient population. Nonetheless, in the future, it will be worthwhile to see how these changes to PCOS diagnosis affect fertility care outcomes.

Whether DOR patients have poorer fertilisation outcomes is similarly controversial, though the literature available to investigate this question is less robust compared to PCOS. Some studies show lower oocyte maturation rates compared to normal ovarian reserve controls,^[15] suggesting that the number and quality of oocytes diminish concurrently. However, our study found no difference in oocyte maturation or fertilisation rate in DOR patients compared to controls. An incremental increase in blastulation rate was noted among DOR patients who underwent minimal stimulation. Because the minimal stimulation protocol relies on a selective oestrogen receptor modulator and a gonadotropin-releasing hormone (GnRH) antagonist to stimulate pituitary production of gonadotropins, it may better replicate physiologic ovarian stimulation, leading to higher-quality oocytes.

This study takes a unique approach to post-retrieval patient counselling and has several strengths. It was performed at a single academic centre over 10 years and focuses on the practical translation of the results findings into a patient counselling tool. We also present a novel finding regarding the potential benefit, rather than simply non-inferiority, of minimal stimulation in patients with DOR.

Limitations to this study include a lack of patient diversity, a retrospective nature, and the use of blastulation, rather than live birth, as the end metric. In addition, this is a single-centre study where doses and protocols may be unique to the institution and limit generalisability. Lastly, limited sample size prevented direct comparisons between DOR cycles that used different protocols, and fold-change calculations were based on comparisons to unexplained infertility rather than direct comparisons between DOR patients, which may be more clinically relevant.

Our results suggest that patients may be counselled that, regardless of ovarian reserve, oocytes can be expected to follow a similar pattern of attrition into the final blastocyst count. Minimal stimulation can be considered for DOR patients with the potential for a higher blastulation rate and lower medication costs.

CONCLUSION

For patients undergoing antagonist protocol IVF cycles, ovarian reserve does not appear to impact embryo attrition rates. However, minimal stimulation protocols utilised for those with DOR had a distinct advantage of a higher blastulation rate compared to a similar patient population undergoing stimulation with high-dose gonadotropins.