Automatic Ploidy Prediction and Quality Assessment of Human Blastocyst Using Time-Lapse Imaging.

Assessing fertilized human embryos is crucial for in vitro-fertilization (IVF), a task being revolutionized by artificial intelligence and deep learning. Existing models used for embryo quality assessment and chromosomal abnormality (ploidy) detection could be significantly improved by effectively utilizing time-lapse imaging to identify critical developmental time points for maximizing prediction accuracy. Addressing this, we developed and compared various embryo ploidy status prediction models across distinct embryo development stages. We present BELA (Blastocyst Evaluation Learning Algorithm), a state-of-the-art ploidy prediction model surpassing previous image- and video-based models, without necessitating subjective input from embryologists. BELA uses multitask learning to predict quality scores that are used downstream to predict ploidy status. By achieving an AUC of 0.76 for discriminating between euploidy and aneuploidy embryos on the Weill Cornell dataset, BELA matches the performance of models trained on embryologists' manual scores. While not a replacement for preimplantation genetic testing for aneuploidy (PGT-A), BELA exemplifies how such models can streamline the embryo evaluation process, reducing time and effort required by embryologists.

A non-invasive artificial intelligence approach for the prediction of human blastocyst ploidy: a retrospective model development and validation study.

BACKGROUND: One challenge in the field of in-vitro fertilisation is the selection of the most viable embryos for transfer. Morphological quality assessment and morphokinetic analysis both have the disadvantage of intra-observer and inter-observer variability. A third method, preimplantation genetic testing for aneuploidy (PGT-A), has limitations too, including its invasiveness and cost. We hypothesised that differences in aneuploid and euploid embryos that allow for model-based classification are reflected in morphology, morphokinetics, and associated clinical information. METHODS: In this retrospective study, we used machine-learning and deep-learning approaches to develop STORK-A, a non-invasive and automated method of embryo evaluation that uses artificial intelligence to predict embryo ploidy status. Our method used a dataset of 10 378 embryos that consisted of static images captured at 110 h after intracytoplasmic sperm injection, morphokinetic parameters, blastocyst morphological assessments, maternal age, and ploidy status. Independent and external datasets, Weill Cornell Medicine EmbryoScope+ (WCM-ES+; Weill Cornell Medicine Center of Reproductive Medicine, NY, USA) and IVI Valencia (IVI Valencia, Health Research Institute la Fe, Valencia, Spain) were used to test the generalisability of STORK-A and were compared measuring accuracy and area under the receiver operating characteristic curve (AUC). FINDINGS: Analysis and model development included the use of 10 378 embryos, all with PGT-A results, from 1385 patients (maternal age range 21-48 years; mean age 36·98 years [SD 4·62]). STORK-A predicted aneuploid versus euploid embryos with an accuracy of 69·3% (95% CI 66·9-71·5; AUC 0·761; positive predictive value [PPV] 76·1%; negative predictive value [NPV] 62·1%) when using images, maternal age, morphokinetics, and blastocyst score. A second classification task trained to predict complex aneuploidy versus euploidy and single aneuploidy produced an accuracy of 74·0% (95% CI 71·7-76·1; AUC 0·760; PPV 54·9%; NPV 87·6%) using an image, maternal age, morphokinetic parameters, and blastocyst grade. A third classification task trained to predict complex aneuploidy versus euploidy had an accuracy of 77·6% (95% CI 75·0-80·0; AUC 0·847; PPV 76·7%; NPV 78·0%). STORK-A reported accuracies of 63·4% (AUC 0·702) on the WCM-ES+ dataset and 65·7% (AUC 0·715) on the IVI Valencia dataset, when using an image, maternal age, and morphokinetic parameters, similar to the STORK-A test dataset accuracy of 67·8% (AUC 0·737), showing generalisability. INTERPRETATION: As a proof of concept, STORK-A shows an ability to predict embryo ploidy in a non-invasive manner and shows future potential as a standardised supplementation to traditional methods of embryo selection and prioritisation for implantation or recommendation for PGT-A. FUNDING: US National Institutes of Health.

Error-prone repair of stalled replication forks drives mutagenesis and loss of heterozygosity in haploinsufficient BRCA1 cells.

Germline mutations in the BRCA genes are associated with a higher risk of carcinogenesis, which is linked to an increased mutation rate and loss of the second unaffected BRCA allele (loss of heterozygosity, LOH). However, the mechanisms triggering mutagenesis are not clearly understood. The BRCA genes contain high numbers of repetitive DNA sequences. We detected replication forks stalling, DNA breaks, and deletions at these sites in haploinsufficient BRCA cells, thus identifying the BRCA genes as fragile sites. Next, we found that stalled forks are repaired by error-prone pathways, such as microhomology-mediated break-induced replication (MMBIR) in haploinsufficient BRCA1 breast epithelial cells. We detected MMBIR mutations in BRCA1 tumor cells and noticed deletions-insertions (>50 bp) at the BRCA1 genes in BRCA1 patients. Altogether, these results suggest that under stress, error-prone repair of stalled forks is upregulated and induces mutations, including complex genomic rearrangements at the BRCA genes (LOH), in haploinsufficient BRCA1 cells.

Blastocyst score, a blastocyst quality ranking tool, is a predictor of blastocyst ploidy and implantation potential.

OBJECTIVE: To convert blastocyst (BL) morphological grade and BL day into a numeric blastocyst score (BS). DESIGN: Retrospective cohort study. SETTING: Academic center. PATIENTS: A total of 5,653 BL of known implantation (fetal heart, FH) and 11,348 biopsied BL. INTERVENTIONS: Based on their FH rates and/or significance, a score (1-4) was assigned to each BL grade component. The BL morphological score (BMS) is the sum (BS = BMS on day 5; BS = BMS + 2 on day 6). MAIN OUTCOME MEASURES: Statistics characterized the FH and euploidy odds with BS. RESULTS: All three morphology grade components and BL day were associated with implantation and euploidy probability. The FH rate and euploidy odds decrease with larger BS. The BS was the most important factor (odds ratio [OR] per unit change = 0.807, 95% confidence interval [CI] 0.784, 0.831) for untested and euploid BL implantation, and the sole one for euploid BL (OR/unit change = 0.845, 95% CI 0.803, 0.889). The BS is the second most significant factor after maternal age for euploidy probability (OR/unit change = 0.808, 95% CI 0.795, 0.822). In training and validation sets (75:25), the BS can predict implantation with similar area under the curve [AUC] (training = 0.628, 95% CI 0.613, 0.643; validation = 0.606, 95% CI 0.581, 0.631). The BS has better euploidy prediction ability (training AUC = 0.683, 95% CI 0.673, 0.693; validation AUC = 0.698, 95% CI 0.681, 0.715). The BS can stratify BL into good (3-5), fair (6-9), and poor (10-14) groups, reflecting their FH, live birth rates, and ploidy status. Advanced maternal age was associated with lower untested BL implantation and lower euploidy odds across all groups. CONCLUSIONS: The BS is a predictor of BL ploidy and FH implantation.

Deep learning enables robust assessment and selection of human blastocysts after in vitro fertilization.

Visual morphology assessment is routinely used for evaluating of embryo quality and selecting human blastocysts for transfer after in vitro fertilization (IVF). However, the assessment produces different results between embryologists and as a result, the success rate of IVF remains low. To overcome uncertainties in embryo quality, multiple embryos are often implanted resulting in undesired multiple pregnancies and complications. Unlike in other imaging fields, human embryology and IVF have not yet leveraged artificial intelligence (AI) for unbiased, automated embryo assessment. We postulated that an AI approach trained on thousands of embryos can reliably predict embryo quality without human intervention. We implemented an AI approach based on deep neural networks (DNNs) to select highest quality embryos using a large collection of human embryo time-lapse images (about 50,000 images) from a high-volume fertility center in the United States. We developed a framework (STORK) based on Google's Inception model. STORK predicts blastocyst quality with an AUC of >0.98 and generalizes well to images from other clinics outside the US and outperforms individual embryologists. Using clinical data for 2182 embryos, we created a decision tree to integrate embryo quality and patient age to identify scenarios associated with pregnancy likelihood. Our analysis shows that the chance of pregnancy based on individual embryos varies from 13.8% (age ≥41 and poor-quality) to 66.3% (age <37 and good-quality) depending on automated blastocyst quality assessment and patient age. In conclusion, our AI-driven approach provides a reproducible way to assess embryo quality and uncovers new, potentially personalized strategies to select embryos.

Blastocyst development rate influences implantation and live birth rates of similarly graded euploid blastocysts.

OBJECTIVE: To determine whether the blastocyst development rate, as assessed by the day of trophectoderm biopsy (day 5 vs. day 6), affects the live birth rate (LBR) of similarly graded euploid blastocysts. DESIGN: Retrospective cohort study. SETTING: Academic medical center. PATIENT(S): Patients who underwent frozen-thawed single euploid blastocyst transfers from 2013 to 2016 were included. Blastocyst morphologic grading was performed on day 5 or day 6 before the biopsy, with embryos designated into the following groups: good (3-6AA, 3-6AB, and 3-6BA), average (2-6BB), and poor (2-6BC and 2-6CB). INTERVENTION(S): Frozen-thawed embryo transfer. MAIN OUTCOME MEASURE(S): Implantation rate (IR) and LBR. RESULT(S): A total of 701 frozen-thawed single euploid blastocyst transfer cycles were included. Cycles in which day 5 blastocysts were transferred (n = 366) were associated with a significantly higher LBR than those in which day 6 blastocysts were transferred (n = 335; 60.4% vs. 44.8%). The odds ratio remained significant after controlling for all confounders, including the blastocyst grading. Furthermore, there was a significant difference in LBRs between good-quality, average-quality, and poor-quality blastocysts (67.8%, 53.4%, and 29.5%, respectively). Embryos reaching good-quality blastocysts on day 5 yielded significantly higher LBR (72.8% vs. 56.5%) and IR (77.7% vs. 58.7%) compared with those reaching similar quality blastocysts on day 6. Similarly, day 5 average-quality embryos conveyed a significantly higher IR compared with day 6 embryos of the same quality (64.4% vs. 53.4%). CONCLUSION(S): In addition to aneuploidy assessment, the speed of embryo development to the blastocyst stage and an evaluation of blastocyst morphology are critical to selecting the best embryo.

Tripolar chromosome segregation drives the association between maternal genotype at variants spanning PLK4 and aneuploidy in human preimplantation embryos.

Aneuploidy is prevalent in human embryos and is the leading cause of pregnancy loss. Many aneuploidies arise during oogenesis, increasing with maternal age. Superimposed on these meiotic aneuploidies are frequent errors occurring during early mitotic divisions, contributing to widespread chromosomal mosaicism. Here we reanalyzed a published dataset comprising preimplantation genetic testing for aneuploidy in 24 653 blastomere biopsies from day-3 cleavage-stage embryos, as well as 17 051 trophectoderm biopsies from day-5 blastocysts. We focused on complex abnormalities that affected multiple chromosomes simultaneously, seeking insights into their formation. In addition to well-described patterns such as triploidy and haploidy, we identified 4.7% of blastomeres possessing characteristic hypodiploid karyotypes. We inferred this signature to have arisen from tripolar chromosome segregation in normally fertilized diploid zygotes or their descendant diploid cells. This could occur via segregation on a tripolar mitotic spindle or by rapid sequential bipolar mitoses without an intervening S-phase. Both models are consistent with time-lapse data from an intersecting set of 77 cleavage-stage embryos, which were enriched for the tripolar signature among embryos exhibiting abnormal cleavage. The tripolar signature was strongly associated with common maternal genetic variants spanning the centrosomal regulator PLK4, driving the association we previously reported with overall mitotic errors. Our findings are consistent with the known capacity of PLK4 to induce tripolar mitosis or precocious M-phase upon dysregulation. Together, our data support tripolar chromosome segregation as a key mechanism generating complex aneuploidy in cleavage-stage embryos and implicate maternal genotype at a quantitative trait locus spanning PLK4 as a factor influencing its occurrence.

Direct Unequal Cleavages: Embryo Developmental Competence, Genetic Constitution and Clinical Outcome.

OBJECTIVE: To investigate the prevalence, developmental potential, chromosomal constitution and clinical outcome of embryos with direct unequal cleavages (DUC). DESIGN: A retrospective observational study. SETTING: Academic Institution. PARTICIPANT: 21,261 embryos from 3,155 cycles cultured in EmbryoScope®. RESULTS: The total incidence of DUCs per embryo occupying the first three cleavages were 26.1%. Depending of the cell stage, DUC rate was 9.8% at first cleavage (DUC-1), 9.1% at second cleavage (DUC-2), and 3.7% at third cleavage (DUC-3) with 3.6% of embryos exhibiting multiple DUCs (DUC-Plus). The occurrence of DUCs was not correlated with female gamete age or source. The incidence of DUC-1 was significantly higher in embryos fertilized by epididymal and testicular sperm (13.6% and 11.4%, respectively) compared to ejaculated sperm (9.1%, all p<0.05). The total incidences of DUCs were strongly correlated with the onset of blastomere multinucleation (MNB) during the first three divisions. In MNB embryos, DUCs incidence are two to three times more likely to develop when compared to non-MNB embryos (OR = 3.11, 95% CI [2.64, 3.67] at 1-cell stage, OR = 2.64, 95% CI [2.39, 2.91] at 2-cell stage and OR = 2.51, 95% CI [1.84, 3.43] at 4-cell stage). The blastocyst formation rates gradually decreased from 61.0% in non-DUC to 40.2% in DUC-3, 18.8% in DUC-2, 8.2% in DUC-1 and 5.6% in multiple DUC embryos (DUC-Plus). The known implantation rates (FH) for day 3 (D3) transfers were 12.42% (n = 3172) in Non-DUC embryos, 6.3% (n = 127) in DUC-3, and 2.7% (n = 260) in DUC-2 embryos. No live births resulted from either DUC-1 (n = 225) or DUC-Plus (n = 100) embryo transfers. For blastocyst transfers, lower implantation rates (33.3%) but similar live birth (LB) rates (40%) were observed if DUC blastocysts were transferred. Comparatively rates in Non-DUC blastocyst were 45.2% and 34.8%, respectively. The euploid rate gradually increased from DUC-1, -2, -3 to Non-DUC (13.3%, 19.5%, 33.3%, 45.6%, p<0.001) for D3 biopsied embryos. Interestingly, the trend of decreased euploidy disappeared in DUC D5/6 biopsied embryos and similar rates were exemplified in DUC (D5 56.3%, D6 35.6%) vs. non-DUC (D5 51.4%, D6 33.8%) embryos. CONCLUSION: Blastocyst formation, implantation potential and euploid rate were significantly reduced in DUC embryos. DUC embryos should be deselected for D3 transfers, but should be culture to blastocyst stage for possible ET.

Cis-acting DNA sequence at a replication origin promotes repeat expansion to fragile X full mutation.

Fragile X syndrome (FXS) is caused by CGG repeat expansion that leads to FMR1 silencing. Women with a premutation allele are at risk of having a full mutation child with FXS. To investigate the mechanism of repeat expansion, we examined the relationship between a single-nucleotide polymorphism (SNP) variant that is linked to repeat expansion in haplogroup D and a replication origin located ∼53 kb upstream of the repeats. This origin is absent in FXS human embryonic stem cells (hESCs), which have the SNP variant C, but present in the nonaffected hESCs, which have a T variant. The SNP maps directly within the replication origin. Interestingly, premutation hESCs have a replication origin and the T variant similar to nonaffected hESCs. These results suggest that a T/C SNP located at a replication origin could contribute to the inactivation of this replication origin in FXS hESCs, leading to altered replication fork progression through the repeats, which could result in repeat expansion to the FXS full mutation.

Derivation of human embryonic stem cells (hESC).

Stem cells are characterized by their absolute or relative lack of specialization their ability for self-renewal, as well as their ability to generate differentiated progeny through cellular lineages with one or more branches. The increased availability of embryonic tissue and greatly improved derivation methods have led to a large increase in the number of hESC lines.

The DNA replication program is altered at the FMR1 locus in fragile X embryonic stem cells.

Fragile X syndrome (FXS) is caused by a CGG repeat expansion in the FMR1 gene that appears to occur during oogenesis and during early embryogenesis. One model proposes that repeat instability depends on the replication fork direction through the repeats such that (CNG)n hairpin-like structures form, causing DNA polymerase to stall and slip. Examining DNA replication fork progression on single DNA molecules at the endogenous FMR1 locus revealed that replication forks stall at CGG repeats in human cells. Furthermore, replication profiles of FXS human embryonic stem cells (hESCs) compared to nonaffected hESCs showed that fork direction through the repeats is altered at the FMR1 locus in FXS hESCs, such that predominantly the CCG strand serves as the lagging-strand template. This is due to the absence of replication initiation that would typically occur upstream of FMR1, suggesting that altered replication origin usage combined with fork stalling promotes repeat instability during early embryonic development.

Human ESC-derived hemogenic endothelial cells undergo distinct waves of endothelial to hematopoietic transition.

UNLABELLED: Several studies have demonstrated that hematopoietic cells originate from endotheliumin early development; however, the phenotypic progression of progenitor cells during human embryonic hemogenesis is not well described. Here, we define the developmental hierarchy among intermediate populations of hematopoietic progenitor cells (HPCs) derived from human embryonic stem cells (hESCs). We genetically modified hESCs to specifically demarcate acquisition of vascular (VE-cadherin) and hematopoietic (CD41a) cell fate and used this dual-reporting transgenic hESC line to observe endothelial to hematopoietic transition by real-time confocal microscopy. Live imaging and clonal analyses revealed a temporal bias in commitment of HPCs that recapitulates discrete waves of lineage differentiation noted during mammalian hemogenesis. Specifically, HPCs isolated at later time points showed reduced capacity to form erythroid/ megakaryocytic cells and exhibited a tendency toward myeloid fate that was enabled by expression of the Notch ligand Dll4 on hESC-derived vascular feeder cells. These data provide a framework for defining HPC lineage potential, elucidate a molecular contribution from the vascular niche in promoting hematopoietic lineage progression, and distinguish unique subpopulations of hemogenic endothelium during hESC differentiation. KEY POINTS: Live imaging of endothelial to hematopoietic conversion identifies distinct subpopulations of hESC-derived hemogenic endothelium. Expression of the Notch ligand DII4 on vascular ECs drives induction of myeloid fate from hESC-derived hematopoietic progenitors.

Lentiviral transduction and clonal selection of hESCs with endothelial-specific transgenic reporters.

Generation of vascular endothelial cells (EC) from human embryonic stem cells (hESC) is a vital component of cell-based strategies for treatment of cardiovascular disease. Before hESC-derived ECs can be administered in therapeutic modalities, however, chemically defined culture conditions must be developed that reproducibly and robustly induce vascular differentiation. One approach to screening for culture conditions that support differentiation of hESCs to any cell type is their genetic modification with exogenous DNA sequence comprising a tissue-specific gene promoter driving reporters such as fluorescent protein or antibiotic drug resistance. The protocols herein provide instructions for the generation of clonal hESC lines containing a reporter transgene that is specifically expressed in vascular endothelial derivatives. Additionally, they demonstrate the methodology employed to assess vascular differentiation from clonal lines. Together, these protocols provide a solid foundation for study of vascular differentiation, and may also be applied, in principle, to studies of other specialized cell types derived from hESCs.

[Outcome of repeated epididymal sperm aspiration or testicular sperm extraction in azoospermic patients].

OBJECTIVE: To review the outcome of repeated percutaneous sperm aspiration (PESA) and testicular sperm extraction (TESE) for intracytoplasmic sperm injection (ICSI). METHODS: Forty-three cycles of 31 cases of azoospermic patients which underwent at least two PESA or TESE for ICSI from January 2001 to December 2002 were collected. The sperm retrieval, fertilization, implantation and clinical pregnancy were analyzed. RESULTS: Twenty-four cases underwent PESA and 7 cases underwent TESE. There were not any complications in these patients. Compared with the first cycle of 154 cases, the fertilization rate, implantation rate and clinical pregnancy rate were 78.39% vs 73.64%, 19.68% vs 18.38% and 34.88% vs 37.91%, respectively(P > 0.05). CONCLUSIONS: Repeated PESA or TESE is safe and well tolerated in azoospermic patients. Compared with the first cycle, the differences of repeated PESA or TESE cycles in fertilization rate, implantation rate and clinical pregnancy rate were not statistically significant.

[Sperm retrieval methods and pregnancy outcome of 100 azoospermia patients].

OBJECTIVES: To review the retrospective treatment results of the azoospermia patients during January 2001 to January 2002 in the fertility center. METHODS: One hundred males attempted intracytoplasmic sperm injection (ICSI) cycle for treatment of azoospermia. All patients were undergone sperm retrieval by percutaneous epididymal sperm aspiration (PESA) or testicular sperm extraction (TESE) while their wives received conventional ovarian hyperstimulation. The hormone levels, testicular histology, the rates of sperm retrieval, fertilization, implantation and pregnancy were analysed and evaluated. RESULTS: Sperm were retrieved by PESA in 76 of 100 (76%) and by TESE in 23 of 100 (23%) men of azoospermia. The fertilization rate, implantation rate and clinical pregnancy rate were 71.3%, 20.35% and 42.11% respectively in PESA group, and 75.18%, 22.05% and 41.60% respectively in TESA group. Thirty-two clinical pregnancies were achieved with 15 ongoing pregnancies and subsequent live delivery for 15 cases in PESA group, and 2 cases of miscarriage, while 10 clinical pregnancies were achieved with 6 ongoing pregnancies, 2 cases of live delivery and 2 cases of miscarriage in TESA group. One case failed to retrieve sperm by TESE and canceled. CONCLUSIONS: Hormonal levels and testicular histology are unable to predict which men with azoospermia will have sperm retrieved by PESA and TESE. PESA and TESE with ICSI are effective methods to treat azoospermia. There were no significant differences in fertilization, implantation and clinical pregnancy rate between two groups.