WISE: whole-scenario embryo identification using self-supervised learning encoder in IVF.

PURPOSE: To study the effectiveness of whole-scenario embryo identification using a self-supervised learning encoder (WISE) in in vitro fertilization (IVF) on time-lapse, cross-device, and cryo-thawed scenarios. METHODS: WISE was based on the vision transformer (ViT) architecture and masked autoencoders (MAE), a self-supervised learning (SSL) method. To train WISE, we prepared three datasets including the SSL pre-training dataset, the time-lapse identification dataset, and the cross-device identification dataset. To identify whether pairs of images were from the same embryos in different scenarios in the downstream identification tasks, embryo images including time-lapse and microscope images were first pre-processed through object detection, cropping, padding, and resizing, and then fed into WISE to get predictions. RESULTS: WISE could accurately identify embryos in the three scenarios. The accuracy was 99.89% on the time-lapse identification dataset, and 83.55% on the cross-device identification dataset. Besides, we subdivided a cryo-thawed evaluation set from the cross-device test set to have a better estimation of how WISE performs in the real-world, and it reached an accuracy of 82.22%. There were approximately 10% improvements in cross-device and cryo-thawed identification tasks after the SSL method was applied. Besides, WISE demonstrated improvements in the accuracy of 9.5%, 12%, and 18% over embryologists in the three scenarios. CONCLUSION: SSL methods can improve embryo identification accuracy even when dealing with cross-device and cryo-thawed paired images. The study is the first to apply SSL in embryo identification, and the results show the promise of WISE for future application in embryo witnessing.

Leveraging federated learning for boosting data privacy and performance in IVF embryo selection.

PURPOSE: To study the effectiveness of federated learning in in vitro fertilization on embryo evaluation tasks. METHODS: This is a retrospective cohort analysis. Two datasets were used in this study. The ploidy status dataset consisted of 10,065 embryo records, 3760 treatments, and 2479 infertile couples from 5 hospitals. The clinical pregnancy dataset consisted of 4495 embryo records, 4495 treatments, and 3704 infertile couples from 4 hospitals. Federated learning and the gradient boosting decision tree algorithm were utilized for modeling. RESULTS: On the ploidy status dataset, the areas under the receiver operating characteristic curves of our model trained with federated learning were 71.78%, 73.10%, 69.39%, 69.72%, and 73.46% for 5 hospitals respectively, showing an average increase of 2.5% compared to those of our model trained without federated learning. On the clinical pregnancy dataset, the areas under the receiver operating characteristic curves of our model trained with federated learning were 72.03%, 56.77%, 61.63%, and 58.58% for 4 hospitals respectively, showing an average increase of 3.08%. CONCLUSIONS: Federated learning can improve data privacy and data security and meanwhile improve the performance of embryo selection tasks by leveraging data from multiple sources. This study demonstrates the effectiveness of federated learning in embryo evaluation, and the results show the promise for future application.

Concordance between different trophectoderm biopsy sites and the inner cell mass of chromosomal composition measured with a next-generation sequencing platform.

STUDY QUESTION: In PGS, does chromosomal constitution differ among trophectoderm (TE) biopsy sites and between them and the inner cell mass (ICM)? SUMMARY ANSWER: The ploidy concordance between ICM and TE was independent of whether the biopsy site in the TE was near to or far from the ICM. WHAT IS KNOWN ALREADY: TE biopsies are considered less harmful to developing embryos than blastomere biopsies. Removal of multi-cellular samples permits high-resolution next-generation sequencing (Veriseq NGS) to detect aneuploidy present in a minority of cells (mosaicism of diploid and aneuploid cells). However, the prevalence of ploidy discrepancies between different TE biopsy sites and the ICM, as well as confined mosaicism (aneuploidy only in a particular area), has not been established. STUDY DESIGN, SIZE, DURATION: Biopsies were taken from a site opposite to the ICM (TE1), near the ICM (TE2) and within the ICM of the same embryo in 33 donated blastocysts obtained from 12 volunteer patients. The samples were analyzed by the Veriseq NGS to assess ploidy concordance. PARTICIPANTS/MATERIALS, SETTING, METHODS: The mean age of the patients was 34.4 years, and samples from all three biopsy sites were achieved in 29 frozen thawed blastocysts. The aneuploid percentage in each sample was interpreted by Veriseq NGS at the finest resolution involving the number of reads after filtering, sample overall noise score, and average quality/alignment scores according to the Veriseq quality control assessment. Ploidy concordance was then assessed between different TE fractions, and between the TE and ICM. MAIN RESULTS AND THE ROLE OF CHANCE: The euploid rates were similar in the TEs and ICM, and no preferential allocation of euploid lineage within a blastocyst was demonstrated. Whether the biopsy site in the TE was near to or far from the ICM, the chromosomal consistency rate was similar [TE1-to-ICM, 86.2% (25/29) versus TE2-to-ICM, 89.7% (26/29); P = 1.0], suggesting that the cells with different chromosomal components may spread randomly throughout the TE. The following two types of inconsistent PGS conclusions between TE and ICM due to confined mosaicism were observed: (i) euploid TE with mosaic ICM (3%) (1/29); and (ii) mosaic TE with euploid ICM (3%) (1/29) or with aneuploid ICM (7%) (2/29). Thus, the overall rate of confined mosaicism was 14% (4/29). LARGE SCALE DATA: N/A. LIMITATION, REASONS FOR CAUTION: The approach used in the present study was affected by biopsy manipulation limitations involving possible cell contamination and the technical challenge of comprehensive chromosomal screening (CCS) procedures. WIDER IMPLICATIONS OF THE FINDINGS: The rate of confined mosaicism in the blastocysts was estimated in this preliminary study, thus, specifying the incidence of biological sampling biases. The results also verified the random distribution of different cell lineages, and the representative value of a single biopsied sample from the TE. STUDY FUNDING AND CONFLICT OF INTEREST(S): No external funding was obtained; all the authors declare no conflicts of interest regarding this study.

Maximum standardised uptake value of positron emission tomography as a predictor of oesophageal cancer outcomes.

OBJECTIVES: This study aimed to analyse the value of pre-operative 18F-fluorodeoxyglucose positron emission tomography (PET)-computed tomography that can predict tumour pathological complete response, tumour histology grade, overall survival, and recurrence-free survival in patients with locally advanced oesophageal squamous cell carcinoma who underwent neoadjuvant chemoradiotherapy (NCRT) followed by surgery. METHODS: We retrospectively reviewed the cases of patients with locally advanced oesophageal squamous cell carcinoma undergoing NCRT followed by surgery. Patients who did not undergo PET within 3 months of surgery were excluded. We set a pre-operative PET maximum standardised uptake value (SUVmax) of > 5 as the threshold and classified the patients into two groups. We analysed the tumour response and histology grade, and compared the overall survival and recurrence-free survival between the two groups. RESULTS: This cohort included 92 patients with oesophageal squamous cell carcinoma who underwent NCRT followed by surgery; 49 patients had a pre-operative PET SUVmax < 5, and 43 patients had a pre-operative PET SUVmax > 5. The patients' pre-operative PET SUVmax correlated with tumour histology, ypT stage, and tumour response. Patients with a pre-operative SUVmax < 5 had better 2-year-overall survival (78% vs. 62%, P < 0.05) and 2-year recurrence-free survival (62% vs. 34%, P < 0.05) than those with a pre-operative SUV > 5. CONCLUSIONS: Pre-operative SUVmax may be useful to predict tumour response, survival, and recurrence in patients with locally advanced oesophageal squamous cell carcinoma who undergo NCRT followed by surgery.

Dependency of mitochondrial quantity on blastocyst timeline obscures its actual effect to pregnancy outcomes.

Objectives: To explore the correlation between mitochondrial quantity and the blastocyst development timeline as well as their respective contributions to early pregnancy. Methods: A retrospective study was conducted using a dataset comprising 2,633 embryos that underwent preimplantation genetic testing for aneuploidy (PGT-A) between January 2016 and December 2023. The study was divided into three subsets to address distinct aspects: the representativeness of a single trophectoderm (TE) biopsy for mitochondrial quantity (n=43), the correlation between morphokinetic features and mitochondrial quantity (n=307), and the association analysis among mitochondrial quantity, blastocyst timeline factor, and reproductive outcomes (n=2,283). Distribution assessment of mitochondrial quantity across an individual blastocyst involved the identification within multiple biopsies and spent culture media. Timeline evaluation included correlating mitochondrial quantity with time-lapse datasets. Finally, multivariate logistic regression models, incorporating potential effectors alongside mitochondrial quantity, were employed to analyze their respective contributions to early pregnancy endpoints. Results: Of distribution assessment, mitochondrial quantity exhibited an even distribution across the entire trophectoderm (Spearman's ρ=0.82), while no detectable mtDNAs in the corresponding spent culture media. Then the timeline correlation study revealed significant association between mitochondrial quantity and blastocyst features of both the day of expanded blastocyst formation (95% Confidence intervals, CIs: 0.27~4.89, p=0.03) and the timing of expanded blastocyst formation (tEB) (95% CIs: -0.24~-0.01, p=0.04) in the regression model, indicating a strong dependency between mitochondrial quantity and the blastocyst development timeline. For the contribution to early pregnancy, multivariate logistic regression models showed that the day of expanded blastocyst formation contributed to four endpoints persistently: positive for HCG (odd ratio, OR: 0.71, p=0.006), gestational sac (OR: 0.78, p=0.04), fetal heartbeat (OR: 0.71, p=0.004), and progression to 14 weeks (OR: 0.69, p=0.002). Contrastingly, no notable correlation was observed between the mitochondrial quantity and these endpoints. Conclusions: Strong interaction was observed between mitochondrial quantity and the blastocyst timeline, particularly the timing of expanded blastocyst formation. It suggests that the primary determinant influencing pregnancy outcomes lies in the time-dependent parameter of blastocyst rather than in the specific mitochondrial quantity.

Similar survival after endoscopic submucosal dissection and esophagectomy in early esophageal cancer and synchronous or metachronous head and neck cancer.

BACKGROUND: Early-stage esophageal cancer is treated using endoscopic submucosal dissection and esophagectomy. Field cancerization in patients with early-stage esophageal cancer affects treatment outcomes and causes synchronous or metachronous head and neck cancers. We hypothesized that esophagectomy could provide better overall and relapse-free survivals in patients with esophageal cancer and synchronous or metachronous head and neck cancer. METHODS: We retrospectively identified patients with early esophageal squamous cell carcinoma and synchronous or metachronous head and neck cancers. We separated the patients into endoscopic submucosal dissection and esophagectomy groups to compare overall and relapse-free survivals. RESULTS: The study included 106 patients, 25 of whom underwent endoscopic submucosal dissection and 81 underwent esophagectomy. Overall and relapse-free survivals did not show significant differences between the two groups for both synchronous and metachronous head and neck cancers. CONCLUSIONS: Endoscopic submucosal dissection could provide similar overall and relapse-free survivals in patients with esophageal cancer and synchronous or metachronous head and neck cancer.

Double lung transplantation is better than single lung transplantation for end-stage chronic obstructive pulmonary disease: a meta-analysis.

BACKGROUND: Lung transplantation is one of the most common treatment options for patients with end-stage chronic obstructive pulmonary disease. However, the choice between single and double lung transplantation for these patients remains a matter of debate. Therefore, we performed a systematic search of medical databases for studies on single lung transplantation, double lung transplantation, and chronic obstructive pulmonary disease. METHODS: The rate ratio and hazard ratio of survival were analyzed. The meta-analysis included 15 case-control and retrospective registry studies. RESULTS: The rate ratios of the 3-year survival (0.937 and P = 0.041) and 5-year survival (0.775 and P = 0.000) were lower for single lung transplantation than for double lung transplantation. However, the hazard ratio did not differ significantly between the two. CONCLUSIONS: Double lung transplantation was found to provide better benefits than single lung transplantation in terms of the long-term survival in patients with chronic obstructive pulmonary disease.

Long-term survival after pulmonary metastasectomy in patients with esophageal squamous cell carcinoma with lung metastasis.

OBJECTIVES: Esophageal squamous cell carcinoma with pulmonary metastasis has a poor prognosis, and the only treatment modality is systemic therapy such as chemotherapy. Previous studies showed that pulmonary metastasectomy may provide benefits and has been suggested in selected patients with colorectal cancer, renal cancer, and sarcoma. However, there were few literatures evaluating the impact and treatment outcome of pulmonary metastasectomy in esophageal squamous cell carcinoma patients with isolated lung metastases. Therefore, we conducted this study. METHODS: We retrospectively reviewed our patients with esophageal squamous cell carcinoma with pulmonary metastasis. Patients with extrapulmonary metastasis were excluded. We categorized them into two groups - the pulmonary resection group and the systemic treatment only group. We compared the overall survival and progression-free survival between groups, and also analyzed the surgical modality, which includes single or multiple port surgery. RESULTS: The analysis included 44 esophageal squamous cell carcinoma patients with lung metastasis. Among these 44 patients, 14 patients have received pulmonary metastasectomy, and 30 patients received systemic treatment only. Patients who received pulmonary metastasectomy had significantly better overall survival (p < 0.0001) and progression-free survival (p = 0.038) than those who received only systemic treatment. The one-year overall survival and progression-free survival were 100% and 48% in patients receiving pulmonary metastatectomy, and 49% and 33% in patients receiving only systemic treatment. Among 14 patients receiving pulmonary metastatectomy, 10 patients underwent single port surgery. There were no postoperative complications in these 14 patients. CONCLUSION: Esophageal squamous cell carcinoma patients with lung metastasis who can receive pulmonary metastasectomy have better prognosis, and some patients can achieve long-term survival. Our findings suggest that aggressive pulmonary metastasectomy is suggested in esophageal squamous cell carcinoma patients with if no contraindication. Key question: How about the role of pulmonary metastasectomy in esophageal squamous cell carcinoma patients with isolated lung metastasis? KEY FINDINGS: Patients who received pulmonary metastasectomy had better overall survival and progression-free survival than those who received only systemic treatment. TAKE HOME MESSAGE: Esophageal cancer with isolated pulmonary metastasis can be treated aggressively with pulmonary metastasectomy if no contraindication.

Reduced mitochondrial DNA content correlate with poor clinical outcomes in cryotransfers with day 6 single euploid embryos.

Objective: To investigate whether the mitochondrial DNA (mtDNA) content of a single biopsy at trophoblast correlates with the developmental potential and reproductive outcomes of blastocyst. Methods: A retrospective analysis applied the dataset of 1,675 embryos with preimplantation genetic testing for aneuploidy (PGT-A) from 1,305 individuals, and 1,383 embryos involved cryotransfers of single euploid embryo between January 2015 and December 2019. The studied cohort was divided for algorithm establishment on the NGS platform (n=40), correlation of biological features (n=1,635), and correlation of reproductive outcomes (n=1,340). Of the algorithm derived from the NGS platform, the reliability and repeatability were validated via qPCR assay and inter-run controls, respectively. Of the correlation across biological features, stratification analyses were applied to evaluate the effect from a single contributor. Eventually, the correlation between the mtDNA ratios and reproductive outcomes was adjusted according to the significant effector(s). Results: The mtDNA ratios showed statistically different between embryos with different days of blastocyst formation ([Day 5]: 1.06 vs. [Day 6]: 0.66, p=0.021), and between embryos with different expansion stages ([Expansion 5]: 1.05 vs. [Expansion 6]: 0.49, p=0.012). None or weakly correlated with the maternal age, morphology, ploidy, and gender. Analyzed by the different days of blastocyst formation with fixed expansion score as 5 in the euploid single embryo transfers (eSET), the day 6 eSET showed significantly lower reduced mtDNA ratio (n=139) in failure groups of fetal heartbeat (p=0.004), ongoing pregnancy (p=0.007), and live birth (p=0.01); however, no correlation between mtDNA ratios and pregnancy outcomes was observed in the day 5 eSET (n=1,201). Conclusions: The study first demonstrated that mtDNA ratio was dependent on the days of blastocyst formation while expansion stage was fixed. Lower mtDNA ratios were observed in the day 6 eSET with adverse outcomes. The present stratification analyses reveal that the timeline of embryo is an important covariate to the mtDNA content.

Improvement of surgical complications using single-lumen endotracheal tube intubation and artificial carbon dioxide pneumothorax in esophagectomy: a meta-analysis.

BACKGROUND: Esophageal cancer has a poor prognosis. Surgery is the main treatment but involves a high risk of complications. Some surgical strategies have tried to eliminate complications. Our meta-analysis tried to find the benefits of single-lumen endotracheal tube intubation with carbon dioxide (CO2) inflation. METHODS: A systematic search of studies on esophagectomy and CO2 inflation was conducted using PubMed, Medline, and Scopus. The odds ratio of post-operative pulmonary complications and anastomosis leakage were the primary outcomes. The standardized mean difference (SMD) in post-operative hospitalization duration was the secondary outcome. RESULTS: The meta-analysis included four case-control studies with a total of 1503 patients. The analysis showed a lower odds ratio of pulmonary complications in the single-lumen endotracheal tube intubation in the CO2 inflation group (odds ratio: 0.756 [95% confidence interval, CI: 0.518 to 1.103]) compared to that in the double-lumen endotracheal tube intubation group, but anastomosis leakage did not improve (odds ratio: 1.056 [95% CI: 0.769 to 1.45])). The SMD in hospitalization duration did not show significant improvement. (SMD: -0.141[95% CI: - 0.248 to - 0.034]). CONCLUSIONS: Single-lumen endotracheal tube intubation with CO2 inflation improved pulmonary complications and shortened the hospitalization duration. However, no benefit in anastomosis leakage was observed.

High concordance in preimplantation genetic testing for aneuploidy between automatic identification via Ion S5 and manual identification via Miseq.

The Ion S5 (Thermo Fisher Scientific) and Miseq (Illumina) NGS systems are both widely used in the clinical laboratories conducting PGT-A. Each system employs discrepant library preparation steps, sequencing principles, and data processing algorithms. The automatic interpretation via Ion Reporter software (Thermo Fisher Scientific) and the manual interpretation via BlueFuse Multi software (Illumina) for chromosomal copy number variation (CNV) represent very different reporting approaches. Thus, it is intriguing to compare their ability of ploidy detection as PGT-A/NGS system. In the present study, four aneuploid cell lines were individually mixed with a diploid cell line at different aneuploid ratios of 0% (0:5), 10% (1:9), 20% (1:4), 40% (2:3), 50% (3:3), 60% (3:2), 80% (4:1) and 100% (5:0) to assess the sensitivity and specificity for whole chromosomal and segmental aneuploidy detection. The clinical biopsies of 107 blastocysts from 46 IVF/PGT-A cycles recruited between December 2019 and February 2020 were used to calculate the concordance. Initially, the pre-amplified products were divided into two aliquots for different library preparation procedures of each system. Applying the same calling criteria, automatic identification was achieved through the Ion Reporter, while well-trained technicians manually identified each sample through the BlueFuse Multi. The results displayed that both systems reliably distinguished chromosomal CNV of the mixtures with at least 10% aneuploidy from karyotypically normal samples ([Ion S5] whole-chromosomal duplication: 2.14 vs. 2.05, p value = 0.009, segmental deletion: 1.88 vs. 2.05, p value = 0.003; [Miseq] whole-chromosomal duplication: 2.12 vs. 2.03, p value = 0.047, segmental deletion: 1.82 vs. 2.03, p value = 0.002). The sensitivity and specificity were comparable between the Ion S5 and Miseq ([sensitivity] 93% vs. 90%, p = 0.78; [specificity] 100% vs. 100%, p value = 1.0). In the 107 clinical biopsies, three displayed chaotic patterns (2.8%), which could not be interpreted for the ploidy. The ploidy concordance was 99.04% (103/104) per embryo and 99.47% (2265/2277) per chromosome pair. Since their ability of detection were proven to be similar, the automatic identification in Ion S5 system presents comparatively faster and more standardized performance.

The Incidence of Mosaicism for Individual Chromosome in Human Blastocysts Is Correlated With Chromosome Length.

Mosaicism, known as partial aneuploidies, mostly originates from mitotic errors during the post-zygotic stage; it consists of different cell lineages within a human embryo. The incidence of mosaicism has not been shown to correlate with maternal age, and its correlation with individual chromosome characteristics has not been well investigated. In this study, the results of preimplantation genetic testing for aneuploidy (PGT-A) derived from 4,036 blastocysts (930 IVF couples) were collected from 2015 to 2017. Via next-generation sequencing for comprehensive chromosome screening, embryo ploidy was identified as aneuploid, mosaic, and euploid. Total mosaicism was classified into two categories: "mosaic euploid/aneuploidy" (with mosaic aneuploidy between 20 and 80%) and "mosaic and aneuploidy" (a uniformly abnormal embryo superimposed with mosaic aneuploidies). Frequency of mosaicism was analyzed according to the function of chromosomal lengths, which divides involved chromosomes into three groups: group A (156-249 Mb), group B (102-145 Mb), and group C (51-90 Mb). The results show that the aneuploidy was more frequent in group C than in group A and group B (A: 23.7%, B: 35.1, 41.2%, p < 0.0001), while the mosaicism was more frequent in group A and group B than in group C [(Mosaic euploid/aneuploid) A: 14.6%, B: 12.4%, C: 9.9%, p < 0.0001; (mosaic and aneuploid) A: 21.3%, B: 22.9%, C: 18.9%, p < 0.0001; (Total mosaicism) A: 35.9%, B: 35.3%, C: 28.8%, p < 0.0001]. The significantly higher frequency of aneuploidy was on the shorter chromosome (< 90 Mb), and that of mosaicism was on the longer chromosomes (> 100 Mb). The length association did not reach significance in the patients with advanced age (≥ 36 years), and of the chromosome-specific mosaicism rate, the highest prevalence was on chromosome 14 (5.8%), 1 (5.7%), and 9 (5.6%). Although the length association was observed via group comparison, there may be affecting mechanisms other than chromosomes length. Eventually, twenty patients with mosaic embryo cryotransfers resulted in six live births. No significant correlation was observed between the transfer outcomes and chromosome length; however, the analysis was limited by small sample size.

Identification of mosaic and segmental aneuploidies by next-generation sequencing in preimplantation genetic screening can improve clinical outcomes compared to array-comparative genomic hybridization.

BACKGROUND: Chromosomal mosaicism is observed as the presence of both euploid and aneuploid cells in a particular blastocyst. Recent studies have reported that the implantation rate of mosaic embryo transfer is remarkably lower than the euploid embryos. The superior capability of next-generation sequencing (NGS) to detect chromosomal mosaicism in preimplantation genetic screening (PGS) remains controversial, and several data displayed similar implantation and pregnancy rates using NGS or array comparative genomic hybridization (aCGH). RESULTS: In this study, the main inconsistency of aneuploidy detection and clinical performance between the NGS and aCGH were assessed. The phase I consisted of a parallel comparison in 182 blastocysts from 45 selected PGS patients for both the NGS and aCGH platforms. The phase II retrospectively compared the clinical outcomes of 90 patients with NGS-screened euploid embryo transfer to that of 129 patients with aCGH-screened euploid embryo transfer. The parallel comparison showed that the inconsistency of embryo euploidy was 11.8% (p = 0.01). Chromosomal mosaicism (10.7% with NGS vs. 3.9% with aCGH) and segmental aneuploidy (10.7% with NGS vs. 6.7% with aCGH) contributed to the discrepancy mainly. The chromosomally mosaic embryos (20%-50% of aneuploidy) and several embryos with segmental aneuploidy (≥10 Mbp) were hard to distinguish using the aCGH platform, but could be clearly identified using the NGS platform. After the first euploid embryo cryotransfer, the ß-HCG(+) rate and implantation rate significantly increased in the PGS/NGS patients (HCG[+] rate: 73.3% in PGS/NGS vs. 60.5% in PGS/aCGH, p = 0.048; implantation rate: 53.2% in PGS/NGS vs. 45.0% in PGS/aCGH, p = 0.043). The clinical and ongoing pregnancy rates appeared higher in the NGS group, but did not reached statistical significance. CONCLUSIONS: The results demonstrated that the NGS platform can identify embryos with chromosomal mosaicism and segmental aneuploidy more precisely than the aCGH platform, and the following clinical performance of NGS was more favorable.

A Patient Friendly Corifollitropin Alfa Protocol without Routine Pituitary Suppression in Normal Responders.

The release of corifollitropin alfa simplifies daily injections of short-acting recombinant follicular stimulating hormone (rFSH), and its widely-used protocol involves short-acting gonadotropins supplements and a fixed GnRH antagonist regimen, largely based on follicle size. In this study, the feasibility of corifollitropin alfa without routine pituitary suppression was evaluated. A total of 288 patients were stimulated by corifollitropin alfa on cycle day 3 following with routine serum hormone monitoring and follicle scanning every other day after 5 days of initial stimulation, and a GnRH antagonist (0.25 mg) was only used prophylactically when the luteinizing hormone (LH) was ≧ 6 IU/L (over half of the definitive LH surge). The incidence of premature LH surge (≧ 10 IU/L) was 2.4% (7/288) before the timely injection of a single GnRH antagonist, and the elevated LH level was dropped down from 11.9 IU/L to 2.2 IU/L after the suppression. Two hundred fifty-one patients did not need any antagonist (87.2% [251/288]) throughout the whole stimulation. No adverse effects were observed regarding oocyte competency (fertilization rate: 78%; blastocyst formation rate: 64%). The live birth rate per OPU cycle after the first cryotransfer was 56.3% (161/286), and the cumulative live birth rate per OPU cycle after cyrotransfers was 69.6% (199/286). Of patients who did and did not receive GnRH antagonist during stimulation, no significant difference existed in the cumulative live birth rates (78.4% vs. 68.3%, p = 0.25). The results demonstrated that the routine GnRH antagonist administration is not required in the corifollitropin-alfa cycles using a flexible and hormone-depended antagonist regimen, while the clinical outcome is not compromised. This finding reveals that the use of a GnRH antagonist only occasionally may be needed.