An artificial intelligence model correlated with morphological and genetic features of blastocyst quality improves ranking of viable embryos.

RESEARCH QUESTION: Can better methods be developed to evaluate the performance and characteristics of an artificial intelligence model for evaluating the likelihood of clinical pregnancy based on analysis of day-5 blastocyst-stage embryos, such that performance evaluation more closely reflects clinical use in IVF procedures, and correlations with known features of embryo quality are identified? DESIGN: De-identified images were provided retrospectively or collected prospectively by IVF clinics using the artificial intelligence model in clinical practice. A total of 9359 images were provided by 18 IVF clinics across six countries, from 4709 women who underwent IVF between 2011 and 2021. Main outcome measures included clinical pregnancy outcome (fetal heartbeat at first ultrasound scan), embryo morphology score, and/or pre-implantation genetic testing for aneuploidy (PGT-A) results. RESULTS: A positive linear correlation of artificial intelligence scores with pregnancy outcomes was found, and up to a 12.2% reduction in time to pregnancy (TTP) was observed when comparing the artificial intelligence model with standard morphological grading methods using a novel simulated cohort ranking method. Artificial intelligence scores were significantly correlated with known morphological features of embryo quality based on the Gardner score, and with previously unknown morphological features associated with embryo ploidy status, including chromosomal abnormalities indicative of severity when considering embryos for transfer during IVF. CONCLUSION: Improved methods for evaluating artificial intelligence for embryo selection were developed, and advantages of the artificial intelligence model over current grading approaches were highlighted, strongly supporting the use of the artificial intelligence model in a clinical setting.

Biosimilar recombinant follitropin alfa preparations versus the reference product (Gonal-F®) in couples undergoing assisted reproductive technology treatment: a systematic review and meta-analysis.

BACKGROUND: Live birth has increasingly been identified as the standard clinical approach to measure the success of medically assisted reproduction (MAR). However, previous analyses comparing biosimilar preparations of follitropin alfa versus the reference product (GONAL-f®, Merck KGaA, Darmstadt, Germany or GONAL-f® RFF; EMD Serono, Inc., Rockland, MA), have had insufficient power to detect differences in clinically meaningful outcomes such as live birth. METHODS: Medline, Embase, the Cochrane Library, Web of Science and clinical trial registries were searched for randomised controlled trials (RCTs) and conference abstracts comparing biosimilar follitropin alfa versus the reference product in controlled ovarian stimulation (COS) cycles published before 31 October 2020. Only studies in humans and publications in English were included. Retrieved studies were screened independently by two authors based on titles and abstracts, and then by full text. INCLUSION CRITERIA: RCTs comparing follitropin alfa biosimilar preparations with the reference product in infertile patients of any age, with any type of infertility for any duration, undergoing COS for the purposes of MAR treatment (including frozen cycles). The primary outcome was live birth. Combined data for biosimilar preparations were analysed using a fixed-effects model. RESULTS: From 292 unique records identified, 17 studies were included in the systematic review, representing five unique RCTs that were included in the meta-analysis. Rates of live birth (RR = 0.83, 95% CI 0.71, 0.97; 4 RCTs, n = 1881, I2 = 0%), clinical pregnancy (RR = 0.82, 95% CI 0.72, 0.94; 4 RCTs, n = 2222, I2 = 0%) and ongoing pregnancy (RR = 0.81, 95% CI 0.68, 0.96; 4 RCTs, n = 1232, I2 = 0%) were significantly lower with biosimilar preparations versus the reference product. Rates of cumulative live birth and cumulative clinical pregnancy were also significantly lower with biosimilars versus the reference product. There was high risk of publication bias. CONCLUSIONS: This meta-analysis included data from RCTs evaluating the efficacy and safety of the biosimilar follitropin alfa preparations and demonstrated lower probability of live birth and pregnancy (ongoing and clinical) in couples treated with biosimilar preparations compared with the reference product. This study provides more insight into the differences between biosimilar r-hFSH preparations and the reference product than previously reported. TRIAL REGISTRATION: Registration number: CRD42019121992 .

Fine-tuning blastocyst selection based on morphology: a multicentre analysis of 2461 single blastocyst transfers.

RESEARCH QUESTION: Which blastocyst morphology parameter is associated with live birth after controlling for female age and endometrial receptivity? DESIGN: Retrospective study including fresh single blastocyst transfers (n = 2461) where the value of serum progesterone on day of human chorionic gonadotrophin trigger (PdHCG) was available. Generalized estimating equation regression models evaluated the independent effects of developmental stage (DevSt), inner cell mass (ICM) and trophectoderm grade on live birth rates while controlling for the confounding effects of female age and PdHCG. RESULTS: DevSt was strongly associated with the probability of live birth (P < 0.0001) independently of female age (odds ratio [OR] 0.89, 95% confidence interval [CI] 0.87-0.91) and PdHCG (OR 0.80, 95% CI 0.74-0.87). For full blastocysts, expanded blastocysts and hatching blastocysts, addition of ICM and trophectoderm grading in the multivariable analysis suggested that besides female age (OR 0.92, 95% CI 0.90-0.94) and PdHCG (OR 0.80, 95% CI 0.73-0.87), only DevSt (P = 0.001) and trophectoderm quality (P = 0.004) were independent predictors of live birth, while the predictive capacity of ICM was no longer significant. The mean probability of live birth was highest for AA blastocysts (35.0%), followed by BA blastocysts (31.2%) and AB blastocysts (27.7%). CONCLUSION: This large study analyses for the first time the independent role of blastocyst morphology in predicting live birth while controlling for female age and PdHCG. Its findings suggest that DevSt and then trophectoderm grade are stronger predictors of live birth over ICM grade when selecting a single blastocyst for transfer.

Inter-observer and intra-observer agreement between embryologists during selection of a single Day 5 embryo for transfer: a multicenter study.

STUDY QUESTION: What is the inter-observer and intra-observer agreement between embryologists when selecting a single Day 5 embryo for transfer? SUMMARY ANSWER: The inter-observer and intra-observer agreement between embryologists when selecting a single Day 5 embryo for transfer was generally good, although not optimal, even among experienced embryologists. WHAT IS KNOWN ALREADY: Previous research on the morphological assessment of early stage (two pronuclei to Day 3) embryos has shown varying levels of inter-observer and intra-observer agreement. However, single blastocyst transfer is now becoming increasingly popular and there are no published data that assess inter-observer and intra-observer agreement when selecting a single embryo for Day 5 transfer. STUDY DESIGN, SIZE, DURATION: This was a prospective study involving 10 embryologists working at five different IVF clinics within a single organization between July 2013 and November 2015. PARTICIPANTS/MATERIALS, SETTING, METHODS: The top 10 embryologists were selected based on their yearly Quality Assurance Program scores for blastocyst grading and were asked to morphologically grade all Day 5 embryos and choose a single embryo for transfer in a survey of 100 cases using 2D images. A total of 1000 decisions were therefore assessed. For each case, Day 5 images were shown, followed by a Day 3 and Day 5 image of the same embryo. Subgroup analyses were also performed based on the following characteristics of embryologists: the level of clinical embryology experience in the laboratory; amount of research experience; number of days per week spent grading embryos. The agreement between these embryologists and the one that scored the embryos on the actual day of transfer was also evaluated. Inter-observer and intra-observer variability was assessed using the kappa coefficient to evaluate the extent of agreement. MAIN RESULTS AND THE ROLE OF CHANCE: This study showed that all 10 embryologists agreed on the embryo chosen for transfer in 50 out of 100 cases. In 93 out of 100 cases, at least 6 out of the 10 embryologists agreed. The inter-observer and intra-observer agreement among embryologists when selecting a single Day 5 embryo for transfer was generally good as assessed by the kappa scores (kappa = 0.734, 95% CI: 0.665-0.791 and 0.759, 95% CI: 0.622-0.833, respectively). The subgroup analyses did not substantially alter the inter-observer and intra-observer agreement among embryologists. The agreement when Day 3 images were included alongside Day 5 images of the same embryos resulted in a change of mind at least three times by each embryologist (on average for <10% of cases) and resulted in a small decrease in inter-observer and intra-observer agreement between embryologists (kappa = 0.676, 95% CI: 0.617-0.724 and 0.752, 95% CI: 0.656-808, respectively).The assessment of the inter-observer agreement with regard to morphological grading of Day 5 embryos showed only a fair-to-moderate agreement, which was observed across all subgroup analyses. The highest overall kappa coefficient was seen for the grading of the developmental stage of an embryo (0.513; 95% CI: 0.492-0.538). The findings were similar when the individual embryologists were compared with the embryologist who made the morphological assessments of the available embryos on the actual day of transfer. LIMITATIONS, REASONS FOR CAUTION: All embryologists had already completed their training and were working under one organization with similar policies between the five clinics. Therefore, the inter-observer agreement might not be as high between embryologists working in clinics with different policies or with different levels of training. WIDER IMPLICATIONS OF THE FINDINGS: The generally good, although not optimal uniformity between participating embryologists when selecting a Day 5 embryo for transfer, as well as, the surprisingly low agreement when morphologically grading Day 5 embryos could be improved, potentially resulting in increased pregnancy rates. Future studies need to be directed toward technologies that can help achieve this. STUDY FUNDING/COMPETING INTERESTS: None declared. TRIAL REGISTRATION NUMBER: Not applicable.

Is early embryo development as observed by time-lapse microscopy dependent on whether fresh or frozen sperm was used for ICSI? A cohort study.

PURPOSE: The aim of this study was to compare timings of key events of embryo development from those originating from either fresh or cryopreserved ejaculate sperm using time-lapse technology. METHODS: In this retrospective observational cohort study, time-lapse technology was used to monitor 1927 embryos from 234 women undergoing intracytoplasmic sperm injection (ICSI) and utilizing either fresh (n = 172 cycles) or cryopreserved ejaculate sperm (n = 62 cycles) for insemination were included in the study. Key developmental events as described in time-lapse were compared with the use of generalized estimating equations (GEE) to adjust for any auto-correlation between the observations. In addition, multivariable logit regression models were used to account for any known baseline differences between the two groups. RESULTS: There were no differences in conventional embryo development such as number of 8-cell embryos by 72 h (p = 0.359), the number of blastocysts by 120 h (p = 0.417), and the number of top quality blastocysts (p = 0.956) between the two groups compared. There were no statistical differences in the timings of any of the key embryo developmental events (PN_t1, NEBD, cytokinesis, t2, t3, t4, t5, t6, t7, t8, tM, tSB, tEB, tHB, s1, s2, s3, cc2, and cc3) when either fresh or cryopreserved ejaculate sperm was used for ICSI. This was also confirmed with conventional morphological assessment. CONCLUSIONS: This observational cohort study has shown that there are no differences in the morphokinetic parameters of early embryo development when either fresh or frozen ejaculate sperm are used for ICSI insemination.

Morphokinetic parameters using time-lapse technology and day 5 embryo quality: a prospective cohort study.

PURPOSE: The aims of this prospective study were to evaluate whether time-lapse parameters can aid in the prediction of day 5 embryo quality and also to assess their discriminatory capacity. METHODS: In this prospective cohort study, we used time-lapse technology to record specific timings of key events for 380 day 5 blastocysts (originating from 108 patients). Generalized estimating equation regression models were used to evaluate the capacity of these markers to identify a top-quality blastocyst. Multivariable regression models were also constructed, aiming to identify the model with the highest capacity to predict a top-quality blastocyst. The discriminatory capacity of single predictors or composite models was assessed with the use of receiver operating characteristic (ROC) analyses. RESULTS: Eight significant predictive parameters of a top-quality blastocyst were identified: s3, t6, t7, t8, tM, tSB, tB and tEB. A ROC analysis of the identified parameters found s3 (area under the curve--AUC 0.585, 95 % CI 0.534-0.635) to have the best individual discriminatory capacity to predict a top-quality blastocyst prior to embryo compaction. The parameter tEB (AUC 0.727, 95 % CI 0.675-0.775) was the best predictor regardless of embryo stage. A model containing s3, t8 and tEB showed a slightly increased discriminatory capacity for top-quality blastocyst prediction (AUC 0.748, 95 % CI 0.697-0.794). CONCLUSIONS: The identified morphokinetic parameters and their cutoffs, albeit of limited clinical value, add to the increasing knowledge concerning the potential predictive markers of a top-quality blastocyst. Additional evidence is necessary before validated time-lapse parameters can be used for embryo selection in IVF laboratories.

Time-lapse algorithms and morphological selection of day-5 embryos for transfer: a preclinical validation study.

OBJECTIVE: To determine the agreement between published time-lapse algorithms in selecting the best day-5 embryo for transfer, as well as the agreement between these algorithms and embryologists. DESIGN: Prospective study. SETTING: Private in vitro fertilization center. PATIENT(S): Four hundred and twenty-eight embryos from 100 cycles cultured in the EmbryoScope. INTERVENTION(S): None. MAIN OUTCOME MEASURE(S): Interalgorithm agreement as assessed by the Fleiss kappa coefficient. RESULT(S): Of seven published algorithms analyzed in this study, only one of the 18 possible pairs showed very good agreement (κ = 0.867); one pair showed good agreement (κ = 0.725), four pairs showed fair agreement (κ = 0.226-0.334), and the remaining 12 pairs showed poor agreement (κ = 0.008-0.149). Even in the best-case scenario, the majority of algorithms showed poor to moderate kappa scores (κ = 0.337-0.722) for the assessment of agreement between the embryo(s) selected as "best" by the algorithms and the embryo that was chosen by the majority (>5) of embryologists, as well as with the embryo that was actually selected in the laboratory on the day of transfer (κ = 0.315-0.802). CONCLUSION(S): The results of this study raise concerns as to whether the tested algorithms are applicable in different clinical settings, emphasizing the need for proper external validation before clinical use.